2007-05-30 21:56:52 +00:00
|
|
|
/* ScummVM - Graphic Adventure Engine
|
|
|
|
*
|
|
|
|
* ScummVM is the legal property of its developers, whose names
|
|
|
|
* are too numerous to list here. Please refer to the COPYRIGHT
|
|
|
|
* file distributed with this source distribution.
|
2003-07-28 01:44:38 +00:00
|
|
|
*
|
2007-05-31 20:28:29 +00:00
|
|
|
* Additional copyright for this file:
|
|
|
|
* Copyright (C) 1994-1998 Revolution Software Ltd.
|
|
|
|
*
|
2003-07-28 01:44:38 +00:00
|
|
|
* This program is free software; you can redistribute it and/or
|
|
|
|
* modify it under the terms of the GNU General Public License
|
|
|
|
* as published by the Free Software Foundation; either version 2
|
|
|
|
* of the License, or (at your option) any later version.
|
|
|
|
*
|
|
|
|
* This program is distributed in the hope that it will be useful,
|
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
|
* GNU General Public License for more details.
|
|
|
|
*
|
|
|
|
* You should have received a copy of the GNU General Public License
|
|
|
|
* along with this program; if not, write to the Free Software
|
2005-10-18 01:30:26 +00:00
|
|
|
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
|
2003-07-28 01:44:38 +00:00
|
|
|
*
|
2006-02-09 15:12:44 +00:00
|
|
|
* $URL$
|
|
|
|
* $Id$
|
2003-07-28 01:44:38 +00:00
|
|
|
*/
|
|
|
|
|
2007-09-19 08:40:12 +00:00
|
|
|
|
2005-01-10 22:06:49 +00:00
|
|
|
#include "common/system.h"
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
#include "common/file.h"
|
2005-02-19 14:02:16 +00:00
|
|
|
|
2003-11-16 14:18:29 +00:00
|
|
|
#include "sword2/sword2.h"
|
2003-10-28 19:51:30 +00:00
|
|
|
#include "sword2/defs.h"
|
2006-02-17 15:07:36 +00:00
|
|
|
#include "sword2/header.h"
|
2006-02-12 19:33:45 +00:00
|
|
|
#include "sword2/screen.h"
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
#include "sword2/console.h"
|
2003-10-28 19:51:30 +00:00
|
|
|
#include "sword2/interpreter.h"
|
2004-02-05 14:19:07 +00:00
|
|
|
#include "sword2/logic.h"
|
|
|
|
#include "sword2/maketext.h"
|
2005-02-19 14:02:16 +00:00
|
|
|
#include "sword2/mouse.h"
|
2004-02-05 14:19:07 +00:00
|
|
|
#include "sword2/resman.h"
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
#include "sword2/router.h"
|
2004-11-14 15:00:01 +00:00
|
|
|
#include "sword2/sound.h"
|
2006-02-09 15:41:23 +00:00
|
|
|
#include "sword2/animation.h"
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2003-10-04 00:52:27 +00:00
|
|
|
namespace Sword2 {
|
|
|
|
|
2003-10-18 08:11:50 +00:00
|
|
|
int32 Logic::fnTestFunction(int32 *params) {
|
|
|
|
// params: 0 address of a flag
|
2003-09-17 13:03:13 +00:00
|
|
|
return IR_CONT;
|
2003-07-28 01:44:38 +00:00
|
|
|
}
|
|
|
|
|
2003-10-18 08:11:50 +00:00
|
|
|
int32 Logic::fnTestFlags(int32 *params) {
|
|
|
|
// params: 0 value of flag
|
2003-09-17 13:03:13 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
2003-07-28 01:44:38 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
int32 Logic::fnRegisterStartPoint(int32 *params) {
|
|
|
|
// params: 0 id of startup script to call - key
|
2008-01-28 00:14:17 +00:00
|
|
|
// 1 pointer to ascii message
|
2003-10-18 08:11:50 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
int32 key = params[0];
|
2005-05-12 13:12:15 +00:00
|
|
|
char *name = (char *)decodePtr(params[1]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
_vm->registerStartPoint(key, name);
|
|
|
|
return IR_CONT;
|
2003-07-28 01:44:38 +00:00
|
|
|
}
|
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
int32 Logic::fnInitBackground(int32 *params) {
|
|
|
|
// this screen defines the size of the back buffer
|
2003-07-28 01:44:38 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// params: 0 res id of normal background layer - cannot be 0
|
|
|
|
// 1 1 yes 0 no for a new palette
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_screen->initBackground(params[0], params[1]);
|
|
|
|
return IR_CONT;
|
2003-07-28 01:44:38 +00:00
|
|
|
}
|
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
/**
|
|
|
|
* This function is used by start scripts.
|
|
|
|
*/
|
2003-07-28 01:44:38 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
int32 Logic::fnSetSession(int32 *params) {
|
|
|
|
// params: 0 id of new run list
|
2003-07-28 01:44:38 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
expressChangeSession(params[0]);
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
2003-07-28 01:44:38 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
int32 Logic::fnBackSprite(int32 *params) {
|
|
|
|
// params: 0 pointer to object's graphic structure
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_router->setSpriteStatus(decodePtr(params[0]), BACK_SPRITE);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnSortSprite(int32 *params) {
|
|
|
|
// params: 0 pointer to object's graphic structure
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_router->setSpriteStatus(decodePtr(params[0]), SORT_SPRITE);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnForeSprite(int32 *params) {
|
|
|
|
// params: 0 pointer to object's graphic structure
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_router->setSpriteStatus(decodePtr(params[0]), FORE_SPRITE);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnRegisterMouse(int32 *params) {
|
|
|
|
// this call would be made from an objects service script 0
|
|
|
|
// the object would be one with no graphic but with a mouse - i.e. a
|
|
|
|
// floor or one whose mouse area is manually defined rather than
|
|
|
|
// intended to fit sprite shape
|
|
|
|
|
|
|
|
// params: 0 pointer to ObjectMouse or 0 for no write to mouse
|
|
|
|
// list
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_vm->_mouse->registerMouse(decodePtr(params[0]), NULL);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnAnim(int32 *params) {
|
|
|
|
// params: 0 pointer to object's logic structure
|
|
|
|
// 1 pointer to object's graphic structure
|
|
|
|
// 2 resource id of animation file
|
|
|
|
|
2005-05-12 06:30:16 +00:00
|
|
|
// Normal forward animation
|
|
|
|
return _router->doAnimate(
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
decodePtr(params[0]),
|
|
|
|
decodePtr(params[1]),
|
2005-05-12 06:30:16 +00:00
|
|
|
params[2], false);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnRandom(int32 *params) {
|
|
|
|
// params: 0 min
|
|
|
|
// 1 max
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
writeVar(RESULT, _vm->_rnd.getRandomNumberRng(params[0], params[1]));
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
2003-07-28 01:44:38 +00:00
|
|
|
}
|
|
|
|
|
2003-10-18 08:11:50 +00:00
|
|
|
int32 Logic::fnPreLoad(int32 *params) {
|
2003-09-17 13:03:13 +00:00
|
|
|
// Forces a resource into memory before it's "officially" opened for
|
|
|
|
// use. eg. if an anim needs to run on smoothly from another,
|
|
|
|
// "preloading" gets it into memory in advance to avoid the cacheing
|
|
|
|
// delay that normally occurs before the first frame.
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2003-10-18 08:11:50 +00:00
|
|
|
// params: 0 resource to preload
|
|
|
|
|
2003-11-16 14:18:29 +00:00
|
|
|
_vm->_resman->openResource(params[0]);
|
|
|
|
_vm->_resman->closeResource(params[0]);
|
2003-10-18 08:11:50 +00:00
|
|
|
return IR_CONT;
|
2003-07-28 01:44:38 +00:00
|
|
|
}
|
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
int32 Logic::fnAddSubject(int32 *params) {
|
|
|
|
// params: 0 id
|
|
|
|
// 1 daves reference number
|
2005-04-24 12:13:03 +00:00
|
|
|
_vm->_mouse->addSubject(params[0], params[1]);
|
2003-09-17 13:03:13 +00:00
|
|
|
return IR_CONT;
|
2003-07-28 01:44:38 +00:00
|
|
|
}
|
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
int32 Logic::fnInteract(int32 *params) {
|
|
|
|
// Run targets action on a subroutine. Called by player on his base
|
|
|
|
// level 0 idle, for example.
|
2003-07-28 01:44:38 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// params: 0 id of target from which we derive action script
|
|
|
|
// reference
|
2003-10-18 08:11:50 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
writeVar(PLAYER_ACTION, 0); // must clear this
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
logicUp((params[0] << 16) | 2); // 3rd script of clicked on id
|
|
|
|
|
|
|
|
// Out, up and around again - pc is saved for current level to be
|
|
|
|
// returned to.
|
|
|
|
return IR_GOSUB;
|
2003-07-28 01:44:38 +00:00
|
|
|
}
|
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
int32 Logic::fnChoose(int32 *params) {
|
|
|
|
// params: none
|
2003-07-28 01:44:38 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// This opcode is used to open the conversation menu. The human is
|
|
|
|
// switched off so there will be no normal mouse engine.
|
2003-10-18 08:11:50 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// The player's choice is piggy-backed on the standard opcode return
|
|
|
|
// values, to be used with the CP_JUMP_ON_RETURNED opcode. As far as I
|
|
|
|
// can tell, this is the only function that uses that feature.
|
|
|
|
|
2005-04-24 12:13:03 +00:00
|
|
|
uint32 response = _vm->_mouse->chooseMouse();
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (response == (uint32)-1)
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_REPEAT;
|
|
|
|
|
2005-04-24 12:13:03 +00:00
|
|
|
return IR_CONT | (response << 3);
|
2003-07-28 01:44:38 +00:00
|
|
|
}
|
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
/**
|
|
|
|
* Walk mega to (x,y,dir). Set RESULT to 0 if it succeeded. Otherwise, set
|
|
|
|
* RESULT to 1.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnWalk(int32 *params) {
|
|
|
|
// params: 0 pointer to object's logic structure
|
|
|
|
// 1 pointer to object's graphic structure
|
|
|
|
// 2 pointer to object's mega structure
|
|
|
|
// 3 pointer to object's walkdata structure
|
|
|
|
// 4 target x-coord
|
|
|
|
// 5 target y-coord
|
|
|
|
// 6 target direction (8 means end walk on ANY direction)
|
|
|
|
|
2005-05-03 09:00:06 +00:00
|
|
|
return _router->doWalk(
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
decodePtr(params[0]),
|
|
|
|
decodePtr(params[1]),
|
|
|
|
decodePtr(params[2]),
|
|
|
|
decodePtr(params[3]),
|
2005-05-03 09:00:06 +00:00
|
|
|
params[4], params[5], params[6]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Walk mega to start position of anim
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnWalkToAnim(int32 *params) {
|
|
|
|
// params: 0 pointer to object's logic structure
|
|
|
|
// 1 pointer to object's graphic structure
|
|
|
|
// 2 pointer to object's mega structure
|
|
|
|
// 3 pointer to object's walkdata structure
|
|
|
|
// 4 anim resource id
|
|
|
|
|
2005-05-03 09:00:06 +00:00
|
|
|
return _router->walkToAnim(
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
decodePtr(params[0]),
|
|
|
|
decodePtr(params[1]),
|
|
|
|
decodePtr(params[2]),
|
|
|
|
decodePtr(params[3]),
|
2005-05-03 09:00:06 +00:00
|
|
|
params[4]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2005-05-03 09:00:06 +00:00
|
|
|
* Turn mega to the specified direction.
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnTurn(int32 *params) {
|
|
|
|
// params: 0 pointer to object's logic structure
|
|
|
|
// 1 pointer to object's graphic structure
|
|
|
|
// 2 pointer to object's mega structure
|
|
|
|
// 3 pointer to object's walkdata structure
|
|
|
|
// 4 target direction
|
|
|
|
|
2005-05-03 09:00:06 +00:00
|
|
|
return _router->doFace(
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
decodePtr(params[0]),
|
|
|
|
decodePtr(params[1]),
|
|
|
|
decodePtr(params[2]),
|
|
|
|
decodePtr(params[3]),
|
2005-05-03 09:00:06 +00:00
|
|
|
params[4]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Stand mega at (x,y,dir)
|
|
|
|
* Sets up the graphic object, but also needs to set the new 'current_dir' in
|
|
|
|
* the mega object, so the router knows in future
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnStandAt(int32 *params) {
|
|
|
|
// params: 0 pointer to object's graphic structure
|
|
|
|
// 1 pointer to object's mega structure
|
|
|
|
// 2 target x-coord
|
|
|
|
// 3 target y-coord
|
|
|
|
// 4 target direction
|
|
|
|
|
2005-05-03 09:00:06 +00:00
|
|
|
_router->standAt(
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
decodePtr(params[0]),
|
|
|
|
decodePtr(params[1]),
|
2005-05-03 09:00:06 +00:00
|
|
|
params[2], params[3], params[4]);
|
2003-09-17 13:03:13 +00:00
|
|
|
return IR_CONT;
|
2003-07-28 01:44:38 +00:00
|
|
|
}
|
2003-09-17 13:03:13 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
/**
|
|
|
|
* Stand mega into the specified direction at current feet coords.
|
2005-05-03 09:00:06 +00:00
|
|
|
* Just needs to call standAt() with current feet coords.
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnStand(int32 *params) {
|
|
|
|
// params: 0 pointer to object's graphic structure
|
|
|
|
// 1 pointer to object's mega structure
|
|
|
|
// 2 target direction
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
byte *ob_mega = decodePtr(params[1]);
|
|
|
|
|
|
|
|
ObjectMega obMega(ob_mega);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
2005-05-03 09:00:06 +00:00
|
|
|
_router->standAt(
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
decodePtr(params[0]),
|
|
|
|
ob_mega, obMega.getFeetX(), obMega.getFeetY(), params[2]);
|
2005-05-03 09:00:06 +00:00
|
|
|
return IR_CONT;
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* stand mega at end position of anim
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnStandAfterAnim(int32 *params) {
|
|
|
|
// params: 0 pointer to object's graphic structure
|
|
|
|
// 1 pointer to object's mega structure
|
|
|
|
// 2 anim resource id
|
|
|
|
|
2005-05-03 09:00:06 +00:00
|
|
|
_router->standAfterAnim(
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
decodePtr(params[0]),
|
|
|
|
decodePtr(params[1]),
|
2005-05-03 09:00:06 +00:00
|
|
|
params[2]);
|
|
|
|
return IR_CONT;
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
2003-10-18 08:11:50 +00:00
|
|
|
int32 Logic::fnPause(int32 *params) {
|
2003-07-28 01:44:38 +00:00
|
|
|
// params: 0 pointer to object's logic structure
|
2003-09-17 13:03:13 +00:00
|
|
|
// 1 number of game-cycles to pause
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2003-09-17 13:03:13 +00:00
|
|
|
// NB. Pause-value of 0 causes script to continue, 1 causes a 1-cycle
|
|
|
|
// quit, 2 gives 2 cycles, etc.
|
2003-07-28 01:44:38 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
ObjectLogic obLogic(decodePtr(params[0]));
|
2003-07-28 01:44:38 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (obLogic.getLooping() == 0) {
|
|
|
|
obLogic.setLooping(1);
|
|
|
|
obLogic.setPause(params[1]);
|
2003-07-28 01:44:38 +00:00
|
|
|
}
|
2003-09-17 13:03:13 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (obLogic.getPause()) {
|
|
|
|
obLogic.setPause(obLogic.getPause() - 1);
|
2003-09-17 13:03:13 +00:00
|
|
|
return IR_REPEAT;
|
2003-07-28 01:44:38 +00:00
|
|
|
}
|
2004-03-30 06:54:27 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obLogic.setLooping(0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnMegaTableAnim(int32 *params) {
|
|
|
|
// params: 0 pointer to object's logic structure
|
|
|
|
// 1 pointer to object's graphic structure
|
|
|
|
// 2 pointer to object's mega structure
|
|
|
|
// 3 pointer to animation table
|
|
|
|
|
2005-05-12 06:30:16 +00:00
|
|
|
// Normal forward anim
|
|
|
|
return _router->megaTableAnimate(
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
decodePtr(params[0]),
|
|
|
|
decodePtr(params[1]),
|
|
|
|
decodePtr(params[2]),
|
|
|
|
decodePtr(params[3]),
|
2005-05-12 06:30:16 +00:00
|
|
|
false);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnAddMenuObject(int32 *params) {
|
|
|
|
// params: 0 pointer to a MenuObject structure to copy down
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_vm->_mouse->addMenuObject(decodePtr(params[0]));
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Start a conversation.
|
|
|
|
*
|
|
|
|
* Note that fnStartConversation() might accidentally be called every time the
|
|
|
|
* script loops back for another chooser, but we only want to reset the chooser
|
|
|
|
* count flag the first time this function is called, i.e. when the talk flag
|
|
|
|
* is zero.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnStartConversation(int32 *params) {
|
|
|
|
// params: none
|
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_mouse->startConversation();
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* End a conversation.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnEndConversation(int32 *params) {
|
|
|
|
// params: none
|
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_mouse->endConversation();
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnSetFrame(int32 *params) {
|
|
|
|
// params: 0 pointer to object's graphic structure
|
|
|
|
// 1 resource id of animation file
|
|
|
|
// 2 frame flag (0=first 1=last)
|
|
|
|
|
|
|
|
int32 res = params[1];
|
|
|
|
assert(res);
|
|
|
|
|
|
|
|
// open the resource (& check it's valid)
|
|
|
|
byte *anim_file = _vm->_resman->openResource(res);
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
assert(_vm->_resman->fetchType(res) == ANIMATION_FILE);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
// set up pointer to the animation header
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
AnimHeader anim_head;
|
|
|
|
|
|
|
|
anim_head.read(_vm->fetchAnimHeader(anim_file));
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
// set up anim resource in graphic object
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
ObjectGraphic obGraph(decodePtr(params[0]));
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obGraph.setAnimResource(res);
|
|
|
|
obGraph.setAnimPc(params[2] ? anim_head.noAnimFrames - 1 : 0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
// Close the anim file and drop out of script
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_vm->_resman->closeResource(obGraph.getAnimResource());
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnRandomPause(int32 *params) {
|
|
|
|
// params: 0 pointer to object's logic structure
|
|
|
|
// 1 minimum number of game-cycles to pause
|
|
|
|
// 2 maximum number of game-cycles to pause
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
ObjectLogic obLogic(decodePtr(params[0]));
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
int32 pars[2];
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (obLogic.getLooping() == 0) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
pars[0] = params[1];
|
|
|
|
pars[1] = params[2];
|
|
|
|
fnRandom(pars);
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
pars[1] = readVar(RESULT);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
pars[0] = params[0];
|
|
|
|
return fnPause(pars);
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnRegisterFrame(int32 *params) {
|
|
|
|
// this call would be made from an objects service script 0
|
|
|
|
|
|
|
|
// params: 0 pointer to mouse structure or NULL for no write to
|
|
|
|
// mouse list (non-zero means write sprite-shape to
|
|
|
|
// mouse list)
|
|
|
|
// 1 pointer to graphic structure
|
|
|
|
// 2 pointer to mega structure or NULL if not a mega
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_vm->_screen->registerFrame(
|
|
|
|
decodePtr(params[0]),
|
|
|
|
decodePtr(params[1]),
|
|
|
|
decodePtr(params[2]));
|
2005-02-19 14:02:16 +00:00
|
|
|
return IR_CONT;
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnNoSprite(int32 *params) {
|
|
|
|
// params: 0 pointer to object's graphic structure
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_router->setSpriteStatus(decodePtr(params[0]), NO_SPRITE);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnSendSync(int32 *params) {
|
|
|
|
// params: 0 sync's recipient
|
|
|
|
// 1 sync value
|
|
|
|
|
2005-11-13 19:29:32 +00:00
|
|
|
sendSync(params[0], params[1]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnUpdatePlayerStats(int32 *params) {
|
|
|
|
// engine needs to know certain info about the player
|
|
|
|
|
|
|
|
// params: 0 pointer to mega structure
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
ObjectMega obMega(decodePtr(params[0]));
|
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
ScreenInfo *screenInfo = _vm->_screen->getScreenInfo();
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
screenInfo->player_feet_x = obMega.getFeetX();
|
|
|
|
screenInfo->player_feet_y = obMega.getFeetY();
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
// for the script
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
writeVar(PLAYER_FEET_X, obMega.getFeetX());
|
|
|
|
writeVar(PLAYER_FEET_Y, obMega.getFeetY());
|
|
|
|
writeVar(PLAYER_CUR_DIR, obMega.getCurDir());
|
|
|
|
writeVar(SCROLL_OFFSET_X, screenInfo->scroll_offset_x);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
debug(5, "fnUpdatePlayerStats: %d %d",
|
|
|
|
obMega.getFeetX(), obMega.getFeetY());
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnPassGraph(int32 *params) {
|
|
|
|
// makes an engine local copy of passed ObjectGraphic - run script 4
|
|
|
|
// of an object to request this used by fnTurnTo(id) etc
|
|
|
|
//
|
|
|
|
// remember, we cannot simply read a compact any longer but instead
|
|
|
|
// must request it from the object itself
|
|
|
|
|
|
|
|
// params: 0 pointer to an ObjectGraphic structure
|
|
|
|
|
|
|
|
warning("fnPassGraph() is a no-op now");
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnInitFloorMouse(int32 *params) {
|
|
|
|
// params: 0 pointer to object's mouse structure
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
byte *ob_mouse = decodePtr(params[0]);
|
2005-02-19 14:02:16 +00:00
|
|
|
ScreenInfo *screenInfo = _vm->_screen->getScreenInfo();
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
// floor is always lowest priority
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
ObjectMouse mouse;
|
|
|
|
|
|
|
|
mouse.x1 = 0;
|
|
|
|
mouse.y1 = 0;
|
|
|
|
mouse.x2 = screenInfo->screen_wide - 1;
|
|
|
|
mouse.y2 = screenInfo->screen_deep - 1;
|
|
|
|
mouse.priority = 9;
|
|
|
|
mouse.pointer = NORMAL_MOUSE_ID;
|
|
|
|
|
|
|
|
mouse.write(ob_mouse);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnPassMega(int32 *params) {
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
// makes an engine local copy of passed mega_structure - run script 4
|
|
|
|
// of an object to request this used by fnTurnTo(id) etc
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
//
|
|
|
|
// remember, we cannot simply read a compact any longer but instead
|
|
|
|
// must request it from the object itself
|
|
|
|
|
2008-01-28 00:14:17 +00:00
|
|
|
// params: 0 pointer to a mega structure
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
memcpy(_engineMega, decodePtr(params[0]), ObjectMega::size());
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Turn mega to face point (x,y) on the floor
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnFaceXY(int32 *params) {
|
|
|
|
// params: 0 pointer to object's logic structure
|
|
|
|
// 1 pointer to object's graphic structure
|
|
|
|
// 2 pointer to object's mega structure
|
|
|
|
// 3 pointer to object's walkdata structure
|
|
|
|
// 4 target x-coord
|
|
|
|
// 5 target y-coord
|
|
|
|
|
2005-05-03 09:00:06 +00:00
|
|
|
return _router->faceXY(
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
decodePtr(params[0]),
|
|
|
|
decodePtr(params[1]),
|
|
|
|
decodePtr(params[2]),
|
|
|
|
decodePtr(params[3]),
|
2005-05-03 09:00:06 +00:00
|
|
|
params[4], params[5]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Causes no more objects in this logic loop to be processed. The logic engine
|
|
|
|
* will restart at the beginning of the new list. The current screen will not
|
|
|
|
* be drawn!
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnEndSession(int32 *params) {
|
|
|
|
// params: 0 id of new run-list
|
|
|
|
|
|
|
|
// terminate current and change to next run-list
|
|
|
|
expressChangeSession(params[0]);
|
|
|
|
|
|
|
|
// stop the script - logic engine will now go around and the new
|
|
|
|
// screen will begin
|
|
|
|
return IR_STOP;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnNoHuman(int32 *params) {
|
|
|
|
// params: none
|
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_mouse->noHuman();
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnAddHuman(int32 *params) {
|
|
|
|
// params: none
|
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_mouse->addHuman();
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Wait for a target to become waiting, i.e. not busy.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnWeWait(int32 *params) {
|
|
|
|
// params: 0 target
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
assert(_vm->_resman->fetchType(params[0]) == GAME_OBJECT);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
// Run the target's get-speech-state script
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
runResScript(params[0], 5);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (readVar(RESULT) == 0) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// The target is busy. Try again.
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_vm->_debugger->_speechScriptWaiting = params[0];
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_REPEAT;
|
|
|
|
}
|
|
|
|
|
|
|
|
// The target is waiting, i.e. not busy.
|
|
|
|
|
|
|
|
_vm->_debugger->_speechScriptWaiting = 0;
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Wait for a target to become waiting, i.e. not busy, send a command to it,
|
|
|
|
* then wait for it to finish.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnTheyDoWeWait(int32 *params) {
|
|
|
|
// params: 0 pointer to ob_logic
|
|
|
|
// 1 target
|
|
|
|
// 2 command
|
|
|
|
// 3 ins1
|
|
|
|
// 4 ins2
|
|
|
|
// 5 ins3
|
|
|
|
// 6 ins4
|
|
|
|
// 7 ins5
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
assert(_vm->_resman->fetchType(params[1]) == GAME_OBJECT);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
// Run the target's get-speech-state script
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
runResScript(params[1], 5);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
ObjectLogic obLogic(decodePtr(params[0]));
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (readVar(RESULT) == 1 && readVar(INS_COMMAND) == 0 && obLogic.getLooping() == 0) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// The target is waiting, i.e. not busy, and there is no other
|
|
|
|
// command queued. We haven't sent the command yet, so do it.
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
debug(5, "fnTheyDoWeWait: sending command to %d", params[1]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_vm->_debugger->_speechScriptWaiting = params[1];
|
|
|
|
obLogic.setLooping(1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
writeVar(SPEECH_ID, params[1]);
|
|
|
|
writeVar(INS_COMMAND, params[2]);
|
|
|
|
writeVar(INS1, params[3]);
|
|
|
|
writeVar(INS2, params[4]);
|
|
|
|
writeVar(INS3, params[5]);
|
|
|
|
writeVar(INS4, params[6]);
|
|
|
|
writeVar(INS5, params[7]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
return IR_REPEAT;
|
|
|
|
}
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (obLogic.getLooping() == 0) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// The command has not been sent yet. Keep waiting.
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_vm->_debugger->_speechScriptWaiting = params[1];
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_REPEAT;
|
|
|
|
}
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (readVar(RESULT) == 0) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// The command has been sent, and the target is busy doing it.
|
|
|
|
// Wait for it to finish.
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
debug(5, "fnTheyDoWeWait: Waiting for %d to finish", params[1]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_vm->_debugger->_speechScriptWaiting = params[1];
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_REPEAT;
|
|
|
|
}
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
debug(5, "fnTheyDoWeWait: %d finished", params[1]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obLogic.setLooping(0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
_vm->_debugger->_speechScriptWaiting = 0;
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Wait for a target to become waiting, i.e. not busy, then send a command to
|
|
|
|
* it.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnTheyDo(int32 *params) {
|
|
|
|
// params: 0 target
|
|
|
|
// 1 command
|
|
|
|
// 2 ins1
|
|
|
|
// 3 ins2
|
|
|
|
// 4 ins3
|
|
|
|
// 5 ins4
|
|
|
|
// 6 ins5
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
assert(_vm->_resman->fetchType(params[0]) == GAME_OBJECT);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
// Run the target's get-speech-state script
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
runResScript(params[0], 5);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (readVar(RESULT) == 1 && !readVar(INS_COMMAND)) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// The target is waiting, i.e. not busy, and there is no other
|
|
|
|
// command queued. Send the command.
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
debug(5, "fnTheyDo: sending command to %d", params[0]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
_vm->_debugger->_speechScriptWaiting = 0;
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
writeVar(SPEECH_ID, params[0]);
|
|
|
|
writeVar(INS_COMMAND, params[1]);
|
|
|
|
writeVar(INS1, params[2]);
|
|
|
|
writeVar(INS2, params[3]);
|
|
|
|
writeVar(INS3, params[4]);
|
|
|
|
writeVar(INS4, params[5]);
|
|
|
|
writeVar(INS5, params[6]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
// The target is busy. Come back again next cycle.
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_vm->_debugger->_speechScriptWaiting = params[0];
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_REPEAT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Route to the left or right hand side of target id, if possible.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnWalkToTalkToMega(int32 *params) {
|
|
|
|
// params: 0 pointer to object's logic structure
|
|
|
|
// 1 pointer to object's graphic structure
|
|
|
|
// 2 pointer to object's mega structure
|
|
|
|
// 3 pointer to object's walkdata structure
|
|
|
|
// 4 id of target mega to face
|
2005-05-03 09:00:06 +00:00
|
|
|
// 5 separation
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
2005-05-03 09:00:06 +00:00
|
|
|
return _router->walkToTalkToMega(
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
decodePtr(params[0]),
|
|
|
|
decodePtr(params[1]),
|
|
|
|
decodePtr(params[2]),
|
|
|
|
decodePtr(params[3]),
|
2005-05-03 09:00:06 +00:00
|
|
|
params[4], params[5]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnFadeDown(int32 *params) {
|
|
|
|
// NONE means up! can only be called when screen is fully faded up -
|
|
|
|
// multiple calls wont have strange effects
|
|
|
|
|
|
|
|
// params: none
|
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
if (_vm->_screen->getFadeStatus() == RDFADE_NONE)
|
|
|
|
_vm->_screen->fadeDown();
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
enum {
|
|
|
|
S_OB_GRAPHIC = 0,
|
|
|
|
S_OB_SPEECH = 1,
|
|
|
|
S_OB_LOGIC = 2,
|
|
|
|
S_OB_MEGA = 3,
|
|
|
|
|
|
|
|
S_TEXT = 4,
|
|
|
|
S_WAV = 5,
|
|
|
|
S_ANIM = 6,
|
|
|
|
S_DIR_TABLE = 7,
|
|
|
|
S_ANIM_MODE = 8
|
|
|
|
};
|
|
|
|
|
|
|
|
/**
|
|
|
|
* It's the super versatile fnSpeak. Text and wavs can be selected in any
|
|
|
|
* combination.
|
|
|
|
*
|
|
|
|
* @note We can assume no human - there should be no human, at least!
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnISpeak(int32 *params) {
|
|
|
|
// params: 0 pointer to ob_graphic
|
|
|
|
// 1 pointer to ob_speech
|
|
|
|
// 2 pointer to ob_logic
|
|
|
|
// 3 pointer to ob_mega
|
|
|
|
// 4 encoded text number
|
|
|
|
// 5 wav res id
|
|
|
|
// 6 anim res id
|
|
|
|
// 7 anim table res id
|
|
|
|
// 8 animation mode 0 lip synced,
|
|
|
|
// 1 just straight animation
|
|
|
|
|
|
|
|
static bool cycle_skip = false;
|
|
|
|
static bool speechRunning;
|
|
|
|
|
|
|
|
// Set up the pointers which we know we'll always need
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
ObjectLogic obLogic(decodePtr(params[S_OB_LOGIC]));
|
|
|
|
ObjectGraphic obGraph(decodePtr(params[S_OB_GRAPHIC]));
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
// FIRST TIME ONLY: create the text, load the wav, set up the anim,
|
|
|
|
// etc.
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (obLogic.getLooping() == 0) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// New fudge to wait for smacker samples to finish
|
|
|
|
// since they can over-run into the game
|
|
|
|
|
|
|
|
if (_vm->_sound->getSpeechStatus() != RDSE_SAMPLEFINISHED)
|
|
|
|
return IR_REPEAT;
|
2005-07-30 21:11:48 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// New fudge for 'fx' subtitles: If subtitles switched off, and
|
|
|
|
// we don't want to use a wav for this line either, then just
|
|
|
|
// quit back to script right now!
|
|
|
|
|
2005-02-20 15:38:48 +00:00
|
|
|
if (!_vm->getSubtitles() && !wantSpeechForLine(params[S_WAV]))
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
|
|
|
|
// Drop out for 1st cycle to allow walks/anims to end and
|
|
|
|
// display last frame before system locks while speech loaded
|
|
|
|
|
|
|
|
if (!cycle_skip) {
|
|
|
|
cycle_skip = true;
|
|
|
|
return IR_REPEAT;
|
|
|
|
}
|
|
|
|
|
|
|
|
cycle_skip = false;
|
|
|
|
|
|
|
|
_vm->_debugger->_textNumber = params[S_TEXT];
|
|
|
|
|
|
|
|
// Pull out the text line to get the official text number
|
|
|
|
// (for wav id). Once the wav id's go into all script text
|
|
|
|
// commands, we'll only need this for debugging.
|
|
|
|
|
|
|
|
uint32 text_res = params[S_TEXT] / SIZE;
|
|
|
|
uint32 local_text = params[S_TEXT] & 0xffff;
|
|
|
|
|
|
|
|
// For testing all text & speech!
|
|
|
|
//
|
|
|
|
// A script loop can send any text number to fnISpeak and it
|
|
|
|
// will only run the valid ones or return with 'result' equal
|
|
|
|
// to '1' or '2' to mean 'invalid text resource' and 'text
|
|
|
|
// number out of range' respectively
|
|
|
|
//
|
|
|
|
// See 'testing_routines' object in George's Player Character
|
|
|
|
// section of linc
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (readVar(SYSTEM_TESTING_TEXT)) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
if (!_vm->_resman->checkValid(text_res)) {
|
|
|
|
// Not a valid resource number - invalid (null
|
|
|
|
// resource)
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
writeVar(RESULT, 1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (_vm->_resman->fetchType(text_res) != TEXT_FILE) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// Invalid - not a text resource
|
|
|
|
_vm->_resman->closeResource(text_res);
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
writeVar(RESULT, 1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (!_vm->checkTextLine(_vm->_resman->openResource(text_res), local_text)) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// Line number out of range
|
|
|
|
_vm->_resman->closeResource(text_res);
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
writeVar(RESULT, 2);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
_vm->_resman->closeResource(text_res);
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
writeVar(RESULT, 0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
byte *text = _vm->fetchTextLine(_vm->_resman->openResource(text_res), local_text);
|
|
|
|
_officialTextNumber = READ_LE_UINT16(text);
|
|
|
|
_vm->_resman->closeResource(text_res);
|
|
|
|
|
|
|
|
// Prevent dud lines from appearing while testing text & speech
|
|
|
|
// since these will not occur in the game anyway
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (readVar(SYSTEM_TESTING_TEXT)) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// If actor number is 0 and text line is just a 'dash'
|
|
|
|
// character
|
|
|
|
if (_officialTextNumber == 0 && text[2] == '-' && text[3] == 0) {
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
writeVar(RESULT, 3);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
// Set the 'looping_flag' and the text-click-delays. We can
|
|
|
|
// left-click past the text after half a second, and
|
|
|
|
// right-click past it after a quarter of a second.
|
2005-07-30 21:11:48 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obLogic.setLooping(1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
_leftClickDelay = 6;
|
|
|
|
_rightClickDelay = 3;
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (readVar(PLAYER_ID) != CUR_PLAYER_ID)
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
debug(5, "(%d) Nico: %s", _officialTextNumber, text + 2);
|
|
|
|
else {
|
2006-07-08 20:45:01 +00:00
|
|
|
debug(5, "(%d) %s: %s", _officialTextNumber, _vm->_resman->fetchName(readVar(ID)), text + 2);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
// Set up the speech animation
|
|
|
|
|
|
|
|
if (params[S_ANIM]) {
|
|
|
|
// Just a straight anim.
|
|
|
|
_animId = params[6];
|
|
|
|
} else if (params[S_DIR_TABLE]) {
|
|
|
|
// Use this direction table to derive the anim
|
|
|
|
// NB. ASSUMES WE HAVE A MEGA OBJECT!!
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
ObjectMega obMega(decodePtr(params[S_OB_MEGA]));
|
|
|
|
byte *anim_table = decodePtr(params[S_DIR_TABLE]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_animId = READ_LE_UINT32(anim_table + 4 * obMega.getCurDir());
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
} else {
|
|
|
|
// No animation choosen
|
|
|
|
_animId = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (_animId) {
|
|
|
|
// Set the talker's graphic to the first frame of this
|
|
|
|
// speech anim for now.
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_speechAnimType = readVar(SPEECHANIMFLAG);
|
|
|
|
obGraph.setAnimResource(_animId);
|
|
|
|
obGraph.setAnimPc(0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
// Default back to looped lip synced anims.
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
writeVar(SPEECHANIMFLAG, 0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
// Set up _textX and _textY for speech panning and/or text
|
|
|
|
// sprite position.
|
|
|
|
|
|
|
|
locateTalker(params);
|
|
|
|
|
2007-12-01 08:15:30 +00:00
|
|
|
// If the speech is associated with a specific animation, and
|
|
|
|
// not just a voice-over, set the focus area to the calculated
|
|
|
|
// position.
|
|
|
|
|
|
|
|
if (_animId) {
|
2007-12-07 20:53:06 +00:00
|
|
|
_vm->_system->setFocusRectangle(Common::Rect::center(_textX, _textY, 192, 128));
|
2007-12-01 08:15:30 +00:00
|
|
|
}
|
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// Is it to be speech or subtitles or both?
|
|
|
|
|
|
|
|
// Assume not running until know otherwise
|
|
|
|
speechRunning = false;
|
|
|
|
|
|
|
|
// New fudge for 'fx' subtitles: If speech is selected, and
|
|
|
|
// this line is allowed speech (not if it's an fx subtitle!)
|
|
|
|
|
|
|
|
if (!_vm->_sound->isSpeechMute() && wantSpeechForLine(_officialTextNumber)) {
|
|
|
|
// If the wavId parameter is zero because not yet
|
|
|
|
// compiled into speech command, we can still get it
|
|
|
|
// from the 1st 2 chars of the text line.
|
|
|
|
|
|
|
|
if (!params[S_WAV])
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
params[S_WAV] = (int32)_officialTextNumber;
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
// Panning goes from -16 (left) to 16 (right)
|
|
|
|
int8 speech_pan = ((_textX - 320) * 16) / 320;
|
|
|
|
|
|
|
|
if (speech_pan < -16)
|
|
|
|
speech_pan = -16;
|
|
|
|
else if (speech_pan > 16)
|
|
|
|
speech_pan = 16;
|
|
|
|
|
|
|
|
uint32 rv = _vm->_sound->playCompSpeech(params[S_WAV], 16, speech_pan);
|
|
|
|
|
|
|
|
if (rv == RD_OK) {
|
|
|
|
// Ok, we've got something to play. Set it
|
|
|
|
// playing now. (We might want to do this the
|
|
|
|
// next cycle, don't know yet.)
|
|
|
|
|
|
|
|
speechRunning = true;
|
|
|
|
_vm->_sound->unpauseSpeech();
|
|
|
|
} else {
|
|
|
|
debug(5, "ERROR: PlayCompSpeech(wav=%d (res=%d pos=%d)) returned %.8x", params[S_WAV], text_res, local_text, rv);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2005-02-20 15:38:48 +00:00
|
|
|
if (_vm->getSubtitles() || !speechRunning) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// We want subtitles, or the speech failed to load.
|
|
|
|
// Either way, we're going to show the text so create
|
|
|
|
// the text sprite.
|
|
|
|
|
|
|
|
formText(params);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
// EVERY TIME: run a cycle of animation, if there is one
|
|
|
|
|
|
|
|
if (_animId) {
|
|
|
|
// There is an animation - Increment the anim frame number.
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obGraph.setAnimPc(obGraph.getAnimPc() + 1);
|
|
|
|
|
|
|
|
byte *anim_file = _vm->_resman->openResource(obGraph.getAnimResource());
|
|
|
|
AnimHeader anim_head;
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
anim_head.read(_vm->fetchAnimHeader(anim_file));
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
if (!_speechAnimType) {
|
|
|
|
// ANIM IS TO BE LIP-SYNC'ED & REPEATING
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (obGraph.getAnimPc() == (int32)anim_head.noAnimFrames) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// End of animation - restart from frame 0
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obGraph.setAnimPc(0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
} else if (speechRunning && _vm->_sound->amISpeaking() == RDSE_QUIET) {
|
|
|
|
// The speech is running, but we're at a quiet
|
|
|
|
// bit. Restart from frame 0 (closed mouth).
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obGraph.setAnimPc(0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
} else {
|
|
|
|
// ANIM IS TO PLAY ONCE ONLY
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (obGraph.getAnimPc() == (int32)anim_head.noAnimFrames - 1) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// Reached the last frame of the anim. Hold
|
|
|
|
// anim on this last frame
|
|
|
|
_animId = 0;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_vm->_resman->closeResource(obGraph.getAnimResource());
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
} else if (_speechAnimType) {
|
|
|
|
// Placed here so we actually display the last frame of the
|
|
|
|
// anim.
|
|
|
|
_speechAnimType = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
// EVERY TIME: FIND OUT IF WE NEED TO STOP THE SPEECH NOW...
|
|
|
|
|
|
|
|
// If there is a wav then we're using that to end the speech naturally
|
|
|
|
|
|
|
|
bool speechFinished = false;
|
|
|
|
|
|
|
|
// If playing a sample
|
|
|
|
|
|
|
|
if (speechRunning) {
|
|
|
|
// Has it finished?
|
|
|
|
if (_vm->_sound->getSpeechStatus() == RDSE_SAMPLEFINISHED)
|
|
|
|
speechFinished = true;
|
|
|
|
} else if (!speechRunning && _speechTime) {
|
|
|
|
// Counting down text time because there is no sample - this
|
|
|
|
// ends the speech
|
|
|
|
|
|
|
|
// if no sample then we're using _speechTime to end speech
|
|
|
|
// naturally
|
|
|
|
|
|
|
|
_speechTime--;
|
|
|
|
if (!_speechTime)
|
|
|
|
speechFinished = true;
|
|
|
|
}
|
|
|
|
|
|
|
|
// Ok, all is running along smoothly - but a click means stop
|
|
|
|
// unnaturally
|
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
int mouseX, mouseY;
|
|
|
|
|
|
|
|
_vm->_mouse->getPos(mouseX, mouseY);
|
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// So that we can go to the options panel while text & speech is
|
|
|
|
// being tested
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (readVar(SYSTEM_TESTING_TEXT) == 0 || mouseY > 0) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
MouseEvent *me = _vm->mouseEvent();
|
|
|
|
|
|
|
|
// Note that we now have TWO click-delays - one for LEFT
|
|
|
|
// button, one for RIGHT BUTTON
|
|
|
|
|
|
|
|
if ((!_leftClickDelay && me && (me->buttons & RD_LEFTBUTTONDOWN)) ||
|
|
|
|
(!_rightClickDelay && me && (me->buttons & RD_RIGHTBUTTONDOWN))) {
|
|
|
|
// Mouse click, after click_delay has expired -> end
|
|
|
|
// the speech.
|
|
|
|
|
|
|
|
// if testing text & speech
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (readVar(SYSTEM_TESTING_TEXT)) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// and RB used to click past text
|
|
|
|
if (me->buttons & RD_RIGHTBUTTONDOWN) {
|
|
|
|
// then we want the previous line again
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
writeVar(SYSTEM_WANT_PREVIOUS_LINE, 1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
} else {
|
|
|
|
// LB just want next line again
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
writeVar(SYSTEM_WANT_PREVIOUS_LINE, 0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
speechFinished = true;
|
|
|
|
|
|
|
|
// if speech sample playing, halt it prematurely
|
|
|
|
if (speechRunning)
|
|
|
|
_vm->_sound->stopSpeech();
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
// If we are finishing the speech this cycle, do the business
|
|
|
|
|
|
|
|
// !speechAnimType, as we want an anim which is playing once to have
|
|
|
|
// finished.
|
|
|
|
|
|
|
|
if (speechFinished && !_speechAnimType) {
|
2007-12-01 08:15:30 +00:00
|
|
|
_vm->_system->clearFocusRectangle();
|
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// If there is text, kill it
|
|
|
|
if (_speechTextBlocNo) {
|
|
|
|
_vm->_fontRenderer->killTextBloc(_speechTextBlocNo);
|
|
|
|
_speechTextBlocNo = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
// if there is a speech anim, end it on closed mouth frame
|
|
|
|
if (_animId) {
|
|
|
|
_animId = 0;
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obGraph.setAnimPc(0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
speechRunning = false;
|
|
|
|
|
|
|
|
// no longer in a script function loop
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obLogic.setLooping(0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
_vm->_debugger->_textNumber = 0;
|
|
|
|
|
|
|
|
// reset to zero, in case text line not even extracted (since
|
|
|
|
// this number comes from the text line)
|
|
|
|
_officialTextNumber = 0;
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
writeVar(RESULT, 0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
// Speech still going, so decrement the click_delay if it's still
|
|
|
|
// active
|
|
|
|
|
|
|
|
if (_leftClickDelay)
|
|
|
|
_leftClickDelay--;
|
|
|
|
|
2008-01-28 00:14:17 +00:00
|
|
|
if (_rightClickDelay)
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
_rightClickDelay--;
|
|
|
|
|
|
|
|
return IR_REPEAT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Reset the object and restart script 1 on level 0
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnTotalRestart(int32 *params) {
|
|
|
|
// mega runs this to restart its base logic again - like being cached
|
|
|
|
// in again
|
|
|
|
|
|
|
|
// params: none
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_curObjectHub.setLogicLevel(0);
|
|
|
|
_curObjectHub.setScriptPc(0, 1);
|
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_TERMINATE;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnSetWalkGrid(int32 *params) {
|
|
|
|
// params: none
|
|
|
|
|
|
|
|
warning("fnSetWalkGrid() is no longer a valid opcode");
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Receive and sequence the commands sent from the conversation script. We have
|
|
|
|
* to do this in a slightly tweeky manner as we can no longer have generic
|
|
|
|
* scripts.
|
|
|
|
*/
|
|
|
|
|
|
|
|
enum {
|
|
|
|
INS_talk = 1,
|
|
|
|
INS_anim = 2,
|
|
|
|
INS_reverse_anim = 3,
|
|
|
|
INS_walk = 4,
|
|
|
|
INS_turn = 5,
|
|
|
|
INS_face = 6,
|
|
|
|
INS_trace = 7,
|
|
|
|
INS_no_sprite = 8,
|
|
|
|
INS_sort = 9,
|
|
|
|
INS_foreground = 10,
|
|
|
|
INS_background = 11,
|
|
|
|
INS_table_anim = 12,
|
|
|
|
INS_reverse_table_anim = 13,
|
|
|
|
INS_walk_to_anim = 14,
|
|
|
|
INS_set_frame = 15,
|
|
|
|
INS_stand_after_anim = 16,
|
|
|
|
INS_quit = 42
|
|
|
|
};
|
|
|
|
|
|
|
|
int32 Logic::fnSpeechProcess(int32 *params) {
|
|
|
|
// params: 0 pointer to ob_graphic
|
|
|
|
// 1 pointer to ob_speech
|
|
|
|
// 2 pointer to ob_logic
|
|
|
|
// 3 pointer to ob_mega
|
|
|
|
// 4 pointer to ob_walkdata
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
ObjectSpeech obSpeech(decodePtr(params[1]));
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
while (1) {
|
|
|
|
int32 pars[9];
|
|
|
|
|
|
|
|
// Check which command we're waiting for, and call the
|
|
|
|
// appropriate function. Once we're done, clear the command
|
|
|
|
// and set wait_state to 1.
|
|
|
|
//
|
|
|
|
// Note: we could save a var and ditch wait_state and check
|
|
|
|
// 'command' for non zero means busy
|
|
|
|
//
|
|
|
|
// Note: I can't see that we ever check the value of wait_state
|
|
|
|
// but perhaps it accesses that memory location directly?
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
switch (obSpeech.getCommand()) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
case 0:
|
|
|
|
break;
|
|
|
|
case INS_talk:
|
|
|
|
pars[0] = params[0]; // ob_graphic
|
|
|
|
pars[1] = params[1]; // ob_speech
|
|
|
|
pars[2] = params[2]; // ob_logic
|
|
|
|
pars[3] = params[3]; // ob_mega
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
pars[4] = obSpeech.getIns1(); // encoded text number
|
|
|
|
pars[5] = obSpeech.getIns2(); // wav res id
|
|
|
|
pars[6] = obSpeech.getIns3(); // anim res id
|
|
|
|
pars[7] = obSpeech.getIns4(); // anim table res id
|
|
|
|
pars[8] = obSpeech.getIns5(); // animation mode - 0 lip synced, 1 just straight animation
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
if (fnISpeak(pars) != IR_REPEAT) {
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obSpeech.setCommand(0);
|
|
|
|
obSpeech.setWaitState(1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
return IR_REPEAT;
|
|
|
|
case INS_turn:
|
|
|
|
pars[0] = params[2]; // ob_logic
|
|
|
|
pars[1] = params[0]; // ob_graphic
|
|
|
|
pars[2] = params[3]; // ob_mega
|
|
|
|
pars[3] = params[4]; // ob_walkdata
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
pars[4] = obSpeech.getIns1(); // direction to turn to
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
if (fnTurn(pars) != IR_REPEAT) {
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obSpeech.setCommand(0);
|
|
|
|
obSpeech.setWaitState(1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
return IR_REPEAT;
|
|
|
|
case INS_face:
|
|
|
|
pars[0] = params[2]; // ob_logic
|
|
|
|
pars[1] = params[0]; // ob_graphic
|
|
|
|
pars[2] = params[3]; // ob_mega
|
|
|
|
pars[3] = params[4]; // ob_walkdata
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
pars[4] = obSpeech.getIns1(); // target
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
if (fnFaceMega(pars) != IR_REPEAT) {
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obSpeech.setCommand(0);
|
|
|
|
obSpeech.setWaitState(1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
return IR_REPEAT;
|
|
|
|
case INS_anim:
|
|
|
|
pars[0] = params[2]; // ob_logic
|
|
|
|
pars[1] = params[0]; // ob_graphic
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
pars[2] = obSpeech.getIns1(); // anim res
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
if (fnAnim(pars) != IR_REPEAT) {
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obSpeech.setCommand(0);
|
|
|
|
obSpeech.setWaitState(1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
return IR_REPEAT;
|
|
|
|
case INS_reverse_anim:
|
|
|
|
pars[0] = params[2]; // ob_logic
|
|
|
|
pars[1] = params[0]; // ob_graphic
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
pars[2] = obSpeech.getIns1(); // anim res
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
if (fnReverseAnim(pars) != IR_REPEAT) {
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obSpeech.setCommand(0);
|
|
|
|
obSpeech.setWaitState(1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
return IR_REPEAT;
|
|
|
|
case INS_table_anim:
|
|
|
|
pars[0] = params[2]; // ob_logic
|
|
|
|
pars[1] = params[0]; // ob_graphic
|
|
|
|
pars[2] = params[3]; // ob_mega
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
pars[3] = obSpeech.getIns1(); // pointer to anim table
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
if (fnMegaTableAnim(pars) != IR_REPEAT) {
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obSpeech.setCommand(0);
|
|
|
|
obSpeech.setWaitState(1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
return IR_REPEAT;
|
|
|
|
case INS_reverse_table_anim:
|
|
|
|
pars[0] = params[2]; // ob_logic
|
|
|
|
pars[1] = params[0]; // ob_graphic
|
|
|
|
pars[2] = params[3]; // ob_mega
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
pars[3] = obSpeech.getIns1(); // pointer to anim table
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
if (fnReverseMegaTableAnim(pars) != IR_REPEAT) {
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obSpeech.setCommand(0);
|
|
|
|
obSpeech.setWaitState(1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
return IR_REPEAT;
|
|
|
|
case INS_no_sprite:
|
|
|
|
fnNoSprite(params); // ob_graphic
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obSpeech.setCommand(0);
|
|
|
|
obSpeech.setWaitState(1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_REPEAT ;
|
|
|
|
case INS_sort:
|
|
|
|
fnSortSprite(params); // ob_graphic
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obSpeech.setCommand(0);
|
|
|
|
obSpeech.setWaitState(1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_REPEAT;
|
|
|
|
case INS_foreground:
|
|
|
|
fnForeSprite(params); // ob_graphic
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obSpeech.setCommand(0);
|
|
|
|
obSpeech.setWaitState(1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_REPEAT;
|
|
|
|
case INS_background:
|
|
|
|
fnBackSprite(params); // ob_graphic
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obSpeech.setCommand(0);
|
|
|
|
obSpeech.setWaitState(1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_REPEAT;
|
|
|
|
case INS_walk:
|
|
|
|
pars[0] = params[2]; // ob_logic
|
|
|
|
pars[1] = params[0]; // ob_graphic
|
|
|
|
pars[2] = params[3]; // ob_mega
|
|
|
|
pars[3] = params[4]; // ob_walkdata
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
pars[4] = obSpeech.getIns1(); // target x
|
|
|
|
pars[5] = obSpeech.getIns2(); // target y
|
|
|
|
pars[6] = obSpeech.getIns3(); // target direction
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
if (fnWalk(pars) != IR_REPEAT) {
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obSpeech.setCommand(0);
|
|
|
|
obSpeech.setWaitState(1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
return IR_REPEAT;
|
|
|
|
case INS_walk_to_anim:
|
|
|
|
pars[0] = params[2]; // ob_logic
|
|
|
|
pars[1] = params[0]; // ob_graphic
|
|
|
|
pars[2] = params[3]; // ob_mega
|
|
|
|
pars[3] = params[4]; // ob_walkdata
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
pars[4] = obSpeech.getIns1(); // anim resource
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
if (fnWalkToAnim(pars) != IR_REPEAT) {
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obSpeech.setCommand(0);
|
|
|
|
obSpeech.setWaitState(1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
return IR_REPEAT;
|
|
|
|
case INS_stand_after_anim:
|
|
|
|
pars[0] = params[0]; // ob_graphic
|
|
|
|
pars[1] = params[3]; // ob_mega
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
pars[2] = obSpeech.getIns1(); // anim resource
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
fnStandAfterAnim(pars);
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obSpeech.setCommand(0);
|
|
|
|
obSpeech.setWaitState(1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_REPEAT;
|
|
|
|
case INS_set_frame:
|
|
|
|
pars[0] = params[0]; // ob_graphic
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
pars[1] = obSpeech.getIns1(); // anim_resource
|
|
|
|
pars[2] = obSpeech.getIns2(); // FIRST_FRAME or LAST_FRAME
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
fnSetFrame(pars);
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obSpeech.setCommand(0);
|
|
|
|
obSpeech.setWaitState(1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_REPEAT;
|
|
|
|
case INS_quit:
|
|
|
|
// That's it - we're finished with this
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obSpeech.setCommand(0);
|
|
|
|
// obSpeech.setWaitState(0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
default:
|
|
|
|
// Unimplemented command - just cancel
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obSpeech.setCommand(0);
|
|
|
|
obSpeech.setWaitState(1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (readVar(SPEECH_ID) == readVar(ID)) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// There's a new command for us! Grab the command -
|
|
|
|
// potentially we only have this cycle to do this - and
|
|
|
|
// set things up so that the command will be picked up
|
|
|
|
// on the next iteration of the while loop.
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
debug(5, "fnSpeechProcess: Received new command %d", readVar(INS_COMMAND));
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
writeVar(SPEECH_ID, 0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obSpeech.setCommand(readVar(INS_COMMAND));
|
|
|
|
obSpeech.setIns1(readVar(INS1));
|
|
|
|
obSpeech.setIns2(readVar(INS2));
|
|
|
|
obSpeech.setIns3(readVar(INS3));
|
|
|
|
obSpeech.setIns4(readVar(INS4));
|
|
|
|
obSpeech.setIns5(readVar(INS5));
|
|
|
|
obSpeech.setWaitState(0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
writeVar(INS_COMMAND, 0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
} else {
|
|
|
|
// No new command. We could run a blink anim (or
|
|
|
|
// something) here.
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obSpeech.setWaitState(1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_REPEAT;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnSetScaling(int32 *params) {
|
|
|
|
// params: 0 pointer to object's mega structure
|
|
|
|
// 1 scale constant A
|
|
|
|
// 2 scale constant B
|
|
|
|
|
|
|
|
// 256 * s = A * y + B
|
|
|
|
|
|
|
|
// Where s is system scale, which itself is (256 * actual_scale) ie.
|
|
|
|
// s == 128 is half size
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
ObjectMega obMega(decodePtr(params[0]));
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obMega.setScaleA(params[1]);
|
|
|
|
obMega.setScaleB(params[2]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnStartEvent(int32 *params) {
|
|
|
|
// params: none
|
|
|
|
|
|
|
|
startEvent();
|
|
|
|
return IR_TERMINATE;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnCheckEventWaiting(int32 *params) {
|
|
|
|
// params: none
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
writeVar(RESULT, checkEventWaiting());
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnRequestSpeech(int32 *params) {
|
|
|
|
// change current script - must be followed by a TERMINATE script
|
|
|
|
// directive
|
|
|
|
|
|
|
|
// params: 0 id of target to catch the event and startup speech
|
|
|
|
// servicing
|
|
|
|
|
|
|
|
// Full script id to interact with - megas run their own 7th script
|
|
|
|
sendEvent(params[0], (params[0] << 16) | 6);
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnGosub(int32 *params) {
|
|
|
|
// params: 0 id of script
|
|
|
|
|
|
|
|
// Hurray, script subroutines. Logic goes up - pc is saved for current
|
|
|
|
// level.
|
|
|
|
logicUp(params[0]);
|
|
|
|
return IR_GOSUB;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Wait for a target to become waiting, i.e. not busy, or until we time out.
|
|
|
|
* This is useful when clicking on a target to talk to it, and it doesn't
|
|
|
|
* reply. This way, we won't lock up.
|
|
|
|
*
|
|
|
|
* If the target becomes waiting, RESULT is set to 0. If we time out, RESULT is
|
|
|
|
* set to 1.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnTimedWait(int32 *params) {
|
|
|
|
// params: 0 ob_logic
|
|
|
|
// 1 target
|
|
|
|
// 2 number of cycles before give up
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
assert(_vm->_resman->fetchType(params[1]) == GAME_OBJECT);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
ObjectLogic obLogic(decodePtr(params[0]));
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (obLogic.getLooping() == 0) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// This is the first time, so set up the time-out.
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obLogic.setLooping(params[2]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
// Run the target's get-speech-state script
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
runResScript(params[1], 5);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (readVar(RESULT) == 1) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// The target is waiting, i.e. not busy
|
|
|
|
|
|
|
|
_vm->_debugger->_speechScriptWaiting = 0;
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obLogic.setLooping(0);
|
|
|
|
writeVar(RESULT, 0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obLogic.setLooping(obLogic.getLooping() - 1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (obLogic.getLooping() == 0) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// Time's up.
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
debug(5, "fnTimedWait: Timed out waiting for %d", params[1]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
_vm->_debugger->_speechScriptWaiting = 0;
|
|
|
|
|
|
|
|
// Clear the event that hasn't been picked up - in theory,
|
|
|
|
// none of this should ever happen.
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
killAllIdsEvents(params[1]);
|
|
|
|
writeVar(RESULT, 1);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
// Target is busy. Keep trying.
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_vm->_debugger->_speechScriptWaiting = params[1];
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_REPEAT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnPlayFx(int32 *params) {
|
|
|
|
// params: 0 sample resource id
|
|
|
|
// 1 type (FX_SPOT, FX_RANDOM, FX_LOOP)
|
|
|
|
// 2 delay (0..65535)
|
|
|
|
// 3 volume (0..16)
|
|
|
|
// 4 pan (-16..16)
|
|
|
|
|
|
|
|
// example script:
|
|
|
|
// fnPlayFx (FXWATER, FX_LOOP, 0, 10, 15);
|
|
|
|
// // fx_water is just a local script flag
|
|
|
|
// fx_water = result;
|
|
|
|
// .
|
|
|
|
// .
|
|
|
|
// .
|
|
|
|
// fnStopFx (fx_water);
|
|
|
|
|
|
|
|
int32 res = params[0];
|
|
|
|
int32 type = params[1];
|
|
|
|
int32 delay = params[2];
|
|
|
|
int32 volume = params[3];
|
|
|
|
int32 pan = params[4];
|
|
|
|
|
|
|
|
_vm->_sound->queueFx(res, type, delay, volume, pan);
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnStopFx(int32 *params) {
|
|
|
|
// params: 0 position in queue
|
|
|
|
if (_vm->_sound->stopFx(params[0]) != RD_OK)
|
|
|
|
debug(5, "SFX ERROR: Trying to stop an inactive sound slot");
|
|
|
|
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Start a tune playing, to play once or to loop until stopped or next one
|
|
|
|
* played.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnPlayMusic(int32 *params) {
|
|
|
|
// params: 0 tune id
|
|
|
|
// 1 loop flag (0 or 1)
|
|
|
|
|
|
|
|
char filename[128];
|
|
|
|
bool loopFlag;
|
|
|
|
uint32 rv;
|
|
|
|
|
|
|
|
loopFlag = (params[1] == FX_LOOP);
|
|
|
|
|
|
|
|
rv = _vm->_sound->streamCompMusic(params[0], loopFlag);
|
|
|
|
|
|
|
|
if (rv)
|
|
|
|
debug(5, "ERROR: streamCompMusic(%s, %d, %d) returned error 0x%.8x", filename, params[0], loopFlag, rv);
|
|
|
|
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnStopMusic(int32 *params) {
|
|
|
|
// params: none
|
|
|
|
|
2005-02-08 08:32:50 +00:00
|
|
|
_vm->_sound->stopMusic(false);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnSetValue(int32 *params) {
|
|
|
|
// temp. function!
|
|
|
|
|
|
|
|
// used for setting far-referenced megaset resource field in mega
|
|
|
|
// object, from start script
|
|
|
|
|
|
|
|
// params: 0 pointer to object's mega structure
|
|
|
|
// 1 value to set it to
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
ObjectMega obMega(decodePtr(params[0]));
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obMega.setMegasetRes(params[1]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnNewScript(int32 *params) {
|
|
|
|
// change current script - must be followed by a TERMINATE script
|
|
|
|
// directive
|
|
|
|
|
|
|
|
// params: 0 id of script
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
writeVar(PLAYER_ACTION, 0); // must clear this
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
logicReplace(params[0]);
|
|
|
|
return IR_TERMINATE;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Like getSync(), but called from scripts. Sets the RESULT variable to
|
|
|
|
* the sync value, or 0 if none is found.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnGetSync(int32 *params) {
|
|
|
|
// params: none
|
|
|
|
|
|
|
|
int slot = getSync();
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
writeVar(RESULT, (slot != -1) ? _syncList[slot].sync : 0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Wait for sync to happen. Sets the RESULT variable to the sync value, once
|
|
|
|
* it has been found.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnWaitSync(int32 *params) {
|
|
|
|
// params: none
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
debug(6, "fnWaitSync: %d waits", readVar(ID));
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
int slot = getSync();
|
|
|
|
|
|
|
|
if (slot == -1)
|
|
|
|
return IR_REPEAT;
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
debug(5, "fnWaitSync: %d got sync %d", readVar(ID), _syncList[slot].sync);
|
|
|
|
writeVar(RESULT, _syncList[slot].sync);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnRegisterWalkGrid(int32 *params) {
|
|
|
|
// params: none
|
|
|
|
|
|
|
|
warning("fnRegisterWalkGrid() is no longer a valid opcode");
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnReverseMegaTableAnim(int32 *params) {
|
|
|
|
// params: 0 pointer to object's logic structure
|
|
|
|
// 1 pointer to object's graphic structure
|
|
|
|
// 2 pointer to object's mega structure
|
|
|
|
// 3 pointer to animation table
|
|
|
|
|
2005-05-12 06:30:16 +00:00
|
|
|
// Reverse anim
|
|
|
|
return _router->megaTableAnimate(
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
decodePtr(params[0]),
|
|
|
|
decodePtr(params[1]),
|
|
|
|
decodePtr(params[2]),
|
|
|
|
decodePtr(params[3]),
|
2005-05-12 06:30:16 +00:00
|
|
|
true);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnReverseAnim(int32 *params) {
|
|
|
|
// params: 0 pointer to object's logic structure
|
|
|
|
// 1 pointer to object's graphic structure
|
|
|
|
// 2 resource id of animation file
|
|
|
|
|
2005-05-12 06:30:16 +00:00
|
|
|
// Reverse anim
|
|
|
|
return _router->doAnimate(
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
decodePtr(params[0]),
|
|
|
|
decodePtr(params[1]),
|
2005-05-12 06:30:16 +00:00
|
|
|
params[2], true);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Mark this object for killing - to be killed when player leaves this screen.
|
|
|
|
* Object reloads and script restarts upon re-entry to screen, which causes
|
|
|
|
* this object's startup logic to be re-run every time we enter the screen.
|
|
|
|
* "Which is nice."
|
|
|
|
*
|
|
|
|
* @note Call ONCE from object's logic script, i.e. in startup code, so not
|
|
|
|
* re-called every time script frops off and restarts!
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnAddToKillList(int32 *params) {
|
|
|
|
// params: none
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
uint32 id = readVar(ID);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
// DON'T EVER KILL GEORGE!
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (id == CUR_PLAYER_ID)
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
|
|
|
|
// Scan the list to see if it's already included
|
|
|
|
|
|
|
|
for (uint32 i = 0; i < _kills; i++) {
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (_objectKillList[i] == id)
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
assert(_kills < OBJECT_KILL_LIST_SIZE); // no room at the inn
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_objectKillList[_kills++] = id;
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
// "another one bites the dust"
|
|
|
|
|
|
|
|
// When we leave the screen, all these object resources are to be
|
|
|
|
// cleaned out of memory and the kill list emptied by doing
|
|
|
|
// '_kills = 0', ensuring that all resources are in fact still in
|
|
|
|
// memory and, more importantly, closed before killing!
|
|
|
|
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Set the standby walk coords to be used by fnWalkToAnim() and
|
|
|
|
* fnStandAfterAnim() when the anim header's start/end coords are zero.
|
|
|
|
* Useful during development; can stay in final game anyway.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnSetStandbyCoords(int32 *params) {
|
|
|
|
// params: 0 x-coord
|
|
|
|
// 1 y-coord
|
|
|
|
// 2 direction (0..7)
|
|
|
|
|
2005-05-03 09:00:06 +00:00
|
|
|
_router->setStandbyCoords(params[0], params[1], params[2]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnBackPar0Sprite(int32 *params) {
|
|
|
|
// params: 0 pointer to object's graphic structure
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_router->setSpriteStatus(decodePtr(params[0]), BGP0_SPRITE);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnBackPar1Sprite(int32 *params) {
|
|
|
|
// params: 0 pointer to object's graphic structure
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_router->setSpriteStatus(decodePtr(params[0]), BGP1_SPRITE);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnForePar0Sprite(int32 *params) {
|
|
|
|
// params: 0 pointer to object's graphic structure
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_router->setSpriteStatus(decodePtr(params[0]), FGP0_SPRITE);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnForePar1Sprite(int32 *params) {
|
|
|
|
// params: 0 pointer to object's graphic structure
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_router->setSpriteStatus(decodePtr(params[0]), FGP1_SPRITE);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnSetPlayerActionEvent(int32 *params) {
|
|
|
|
// we want to intercept the player character and have him interact
|
|
|
|
// with an object - from script this code is the same as the mouse
|
|
|
|
// engine calls when you click on an object - here, a third party
|
|
|
|
// does the clicking IYSWIM
|
|
|
|
|
|
|
|
// note - this routine used CUR_PLAYER_ID as the target
|
|
|
|
|
|
|
|
// params: 0 id to interact with
|
|
|
|
|
|
|
|
setPlayerActionEvent(CUR_PLAYER_ID, params[0]);
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Set the special scroll offset variables
|
|
|
|
*
|
|
|
|
* Call when starting screens and to change the camera within screens
|
|
|
|
*
|
|
|
|
* call AFTER fnInitBackground() to override the defaults
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnSetScrollCoordinate(int32 *params) {
|
|
|
|
// params: 0 feet_x value
|
2008-01-28 00:14:17 +00:00
|
|
|
// 1 feet_y value
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
2008-01-28 00:14:17 +00:00
|
|
|
// Called feet_x and feet_y to retain intellectual compatibility with
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// Sword1!
|
|
|
|
//
|
|
|
|
// feet_x & feet_y refer to the physical screen coords where the
|
|
|
|
// system will try to maintain George's feet
|
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
ScreenInfo *screenInfo = _vm->_screen->getScreenInfo();
|
|
|
|
|
|
|
|
screenInfo->feet_x = params[0];
|
|
|
|
screenInfo->feet_y = params[1];
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Stand mega at start position of anim
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnStandAtAnim(int32 *params) {
|
|
|
|
// params: 0 pointer to object's graphic structure
|
|
|
|
// 1 pointer to object's mega structure
|
|
|
|
// 2 anim resource id
|
|
|
|
|
2005-05-03 09:00:06 +00:00
|
|
|
_router->standAtAnim(
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
decodePtr(params[0]),
|
|
|
|
decodePtr(params[1]),
|
2005-05-03 09:00:06 +00:00
|
|
|
params[2]);
|
|
|
|
return IR_CONT;
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
#define SCROLL_MOUSE_WIDTH 20
|
|
|
|
|
|
|
|
int32 Logic::fnSetScrollLeftMouse(int32 *params) {
|
|
|
|
// params: 0 pointer to object's mouse structure
|
|
|
|
|
2008-01-28 00:14:17 +00:00
|
|
|
byte *ob_mouse = decodePtr(params[0]);
|
2005-02-19 14:02:16 +00:00
|
|
|
ScreenInfo *screenInfo = _vm->_screen->getScreenInfo();
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
// Highest priority
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
ObjectMouse mouse;
|
|
|
|
|
|
|
|
mouse.x1 = 0;
|
|
|
|
mouse.y1 = 0;
|
|
|
|
mouse.x2 = screenInfo->scroll_offset_x + SCROLL_MOUSE_WIDTH;
|
|
|
|
mouse.y2 = screenInfo->screen_deep - 1;
|
|
|
|
mouse.priority = 0;
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
if (screenInfo->scroll_offset_x > 0) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// not fully scrolled to the left
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
mouse.pointer = SCROLL_LEFT_MOUSE_ID;
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
} else {
|
|
|
|
// so the mouse area doesn't get registered
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
mouse.pointer = 0;
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
mouse.write(ob_mouse);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnSetScrollRightMouse(int32 *params) {
|
|
|
|
// params: 0 pointer to object's mouse structure
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
byte *ob_mouse = decodePtr(params[0]);
|
2005-02-19 14:02:16 +00:00
|
|
|
ScreenInfo *screenInfo = _vm->_screen->getScreenInfo();
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
// Highest priority
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
ObjectMouse mouse;
|
|
|
|
|
|
|
|
mouse.x1 = screenInfo->scroll_offset_x + _vm->_screen->getScreenWide() - SCROLL_MOUSE_WIDTH;
|
|
|
|
mouse.y1 = 0;
|
|
|
|
mouse.x2 = screenInfo->screen_wide - 1;
|
|
|
|
mouse.y2 = screenInfo->screen_deep - 1;
|
|
|
|
mouse.priority = 0;
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
if (screenInfo->scroll_offset_x < screenInfo->max_scroll_offset_x) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// not fully scrolled to the right
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
mouse.pointer = SCROLL_RIGHT_MOUSE_ID;
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
} else {
|
|
|
|
// so the mouse area doesn't get registered
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
mouse.pointer = 0;
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
mouse.write(ob_mouse);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnColour(int32 *params) {
|
|
|
|
// set border colour - useful during script development
|
|
|
|
// eg. set to colour during a timer situation, then black when timed
|
|
|
|
// out
|
|
|
|
|
|
|
|
// params 0: colour (see defines above)
|
|
|
|
|
|
|
|
#ifdef SWORD2_DEBUG
|
|
|
|
// what colour?
|
|
|
|
switch (params[0]) {
|
|
|
|
case BLACK:
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_screen->setPalette(0, 1, black, RDPAL_INSTANT);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
break;
|
|
|
|
case WHITE:
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_screen->setPalette(0, 1, white, RDPAL_INSTANT);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
break;
|
|
|
|
case RED:
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_screen->setPalette(0, 1, red, RDPAL_INSTANT);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
break;
|
|
|
|
case GREEN:
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_screen->setPalette(0, 1, green, RDPAL_INSTANT);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
break;
|
|
|
|
case BLUE:
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_screen->setPalette(0, 1, blue, RDPAL_INSTANT);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
#ifdef SWORD2_DEBUG
|
|
|
|
#define BLACK 0
|
|
|
|
#define WHITE 1
|
|
|
|
#define RED 2
|
|
|
|
#define GREEN 3
|
|
|
|
#define BLUE 4
|
|
|
|
|
2006-10-15 02:15:38 +00:00
|
|
|
static const uint8 black[4] = { 0, 0, 0, 0 };
|
|
|
|
static const uint8 white[4] = { 255, 255, 255, 0 };
|
|
|
|
static const uint8 red[4] = { 255, 0, 0, 0 };
|
|
|
|
static const uint8 green[4] = { 0, 255, 0, 0 };
|
|
|
|
static const uint8 blue[4] = { 0, 0, 255, 0 };
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
#endif
|
|
|
|
|
|
|
|
int32 Logic::fnFlash(int32 *params) {
|
|
|
|
// flash colour 0 (ie. border) - useful during script development
|
|
|
|
// eg. fnFlash(BLUE) where a text line is missed; RED when some code
|
|
|
|
// missing, etc
|
|
|
|
|
|
|
|
// params: 0 colour to flash
|
|
|
|
|
|
|
|
#ifdef SWORD2_DEBUG
|
|
|
|
// what colour?
|
|
|
|
switch (params[0]) {
|
|
|
|
case WHITE:
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_screen->setPalette(0, 1, white, RDPAL_INSTANT);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
break;
|
|
|
|
case RED:
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_screen->setPalette(0, 1, red, RDPAL_INSTANT);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
break;
|
|
|
|
case GREEN:
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_screen->setPalette(0, 1, green, RDPAL_INSTANT);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
break;
|
|
|
|
case BLUE:
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_screen->setPalette(0, 1, blue, RDPAL_INSTANT);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
// There used to be a busy-wait loop here, so I don't know how long
|
|
|
|
// the delay was meant to be. Probably doesn't matter much.
|
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_screen->updateDisplay();
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
_vm->_system->delayMillis(250);
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_screen->setPalette(0, 1, black, RDPAL_INSTANT);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
#endif
|
|
|
|
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnPreFetch(int32 *params) {
|
|
|
|
// Go fetch resource in the background.
|
|
|
|
|
|
|
|
// params: 0 resource to fetch [guess]
|
|
|
|
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Reverse of fnPassPlayerSaveData() - run script 8 of player object.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnGetPlayerSaveData(int32 *params) {
|
|
|
|
// params: 0 pointer to object's logic structure
|
|
|
|
// 1 pointer to object's graphic structure
|
|
|
|
// 2 pointer to object's mega structure
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
byte *ob_logic = decodePtr(params[0]);
|
|
|
|
byte *ob_graph = decodePtr(params[1]);
|
|
|
|
byte *ob_mega = decodePtr(params[2]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
// Copy from savegame buffers to player object
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
memcpy(ob_logic, _saveLogic, ObjectLogic::size());
|
|
|
|
memcpy(ob_graph, _saveGraphic, ObjectGraphic::size());
|
|
|
|
memcpy(ob_mega, _saveMega, ObjectMega::size());
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
2008-01-28 00:14:17 +00:00
|
|
|
// Any walk-data must be cleared - the player will be set to stand if
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// he was walking when saved.
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
ObjectMega obMega(ob_mega);
|
|
|
|
|
|
|
|
if (obMega.getIsWalking()) {
|
|
|
|
ObjectLogic obLogic(ob_logic);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obMega.setIsWalking(0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
int32 pars[3];
|
|
|
|
|
|
|
|
pars[0] = params[1]; // ob_graphic;
|
|
|
|
pars[1] = params[2]; // ob_mega
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
pars[2] = obMega.getCurDir();
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
fnStand(pars);
|
|
|
|
|
|
|
|
// Reset looping flag (which would have been 1 during fnWalk)
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obLogic.setLooping(0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Copies the 4 essential player structures into the savegame header - run
|
|
|
|
* script 7 of player object to request this.
|
|
|
|
*
|
|
|
|
* Remember, we cannot simply read a compact any longer but instead must
|
|
|
|
* request it from the object itself.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnPassPlayerSaveData(int32 *params) {
|
|
|
|
// params: 0 pointer to object's logic structure
|
|
|
|
// 1 pointer to object's graphic structure
|
|
|
|
// 2 pointer to object's mega structure
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
// Copy from player object to savegame buffers
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
memcpy(_saveLogic, decodePtr(params[0]), ObjectLogic::size());
|
|
|
|
memcpy(_saveGraphic, decodePtr(params[1]), ObjectGraphic::size());
|
|
|
|
memcpy(_saveMega, decodePtr(params[2]), ObjectMega::size());
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnSendEvent(int32 *params) {
|
|
|
|
// we want to intercept the player character and have him interact
|
|
|
|
// with an object - from script
|
|
|
|
|
|
|
|
// params: 0 id to receive event
|
|
|
|
// 1 script to run
|
|
|
|
|
|
|
|
sendEvent(params[0], params[1]);
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Add this walkgrid resource to the list of those used for routing in this
|
|
|
|
* location. Note that this is ignored if the resource is already in the list.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnAddWalkGrid(int32 *params) {
|
|
|
|
// params: 0 id of walkgrid resource
|
|
|
|
|
|
|
|
// All objects that add walkgrids must be restarted whenever we
|
|
|
|
// re-enter a location.
|
|
|
|
|
|
|
|
// DON'T EVER KILL GEORGE!
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (readVar(ID) != CUR_PLAYER_ID) {
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// Need to call this in case it wasn't called in script!
|
|
|
|
fnAddToKillList(NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
_router->addWalkGrid(params[0]);
|
|
|
|
fnPreLoad(params);
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Remove this walkgrid resource from the list of those used for routing in
|
|
|
|
* this location. Note that this is ignored if the resource isn't actually
|
|
|
|
* in the list.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnRemoveWalkGrid(int32 *params) {
|
|
|
|
// params: 0 id of walkgrid resource
|
|
|
|
|
|
|
|
_router->removeWalkGrid(params[0]);
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
// like fnCheckEventWaiting, but starts the event rather than setting RESULT
|
|
|
|
// to 1
|
|
|
|
|
|
|
|
int32 Logic::fnCheckForEvent(int32 *params) {
|
|
|
|
// params: none
|
|
|
|
|
|
|
|
if (checkEventWaiting()) {
|
|
|
|
startEvent();
|
|
|
|
return IR_TERMINATE;
|
|
|
|
}
|
|
|
|
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
// combination of fnPause and fnCheckForEvent
|
|
|
|
// - ie. does a pause, but also checks for event each cycle
|
|
|
|
|
|
|
|
int32 Logic::fnPauseForEvent(int32 *params) {
|
|
|
|
// params: 0 pointer to object's logic structure
|
|
|
|
// 1 number of game-cycles to pause
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
ObjectLogic obLogic(decodePtr(params[0]));
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
if (checkEventWaiting()) {
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
obLogic.setLooping(0);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
startEvent();
|
|
|
|
return IR_TERMINATE;
|
|
|
|
}
|
|
|
|
|
|
|
|
return fnPause(params);
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnClearEvent(int32 *params) {
|
|
|
|
// params: none
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
clearEvent(readVar(ID));
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnFaceMega(int32 *params) {
|
|
|
|
// params: 0 pointer to object's logic structure
|
|
|
|
// 1 pointer to object's graphic structure
|
|
|
|
// 2 pointer to object's mega structure
|
|
|
|
// 3 pointer to object's walkdata structure
|
|
|
|
// 4 id of target mega to face
|
|
|
|
|
2005-05-03 09:00:06 +00:00
|
|
|
return _router->faceMega(
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
decodePtr(params[0]),
|
|
|
|
decodePtr(params[1]),
|
|
|
|
decodePtr(params[2]),
|
|
|
|
decodePtr(params[3]),
|
2005-05-03 09:00:06 +00:00
|
|
|
params[4]);
|
2003-07-28 01:44:38 +00:00
|
|
|
}
|
2003-09-17 13:03:13 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
int32 Logic::fnPlaySequence(int32 *params) {
|
|
|
|
// params: 0 pointer to null-terminated ascii filename
|
2008-01-28 00:14:17 +00:00
|
|
|
// 1 number of frames in the sequence, used for PSX.
|
2003-07-28 01:44:38 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
char filename[30];
|
2003-07-28 01:44:38 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// The original code had some #ifdef blocks for skipping or muting the
|
|
|
|
// cutscenes - fondly described as "the biggest fudge in the history
|
|
|
|
// of computer games" - but at the very least we want to show the
|
|
|
|
// cutscene subtitles, so I removed them.
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2005-05-12 13:12:15 +00:00
|
|
|
debug(5, "fnPlaySequence(\"%s\");", (const char *)decodePtr(params[0]));
|
2003-07-28 01:44:38 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// add the appropriate file extension & play it
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2005-05-12 13:12:15 +00:00
|
|
|
strcpy(filename, (const char *)decodePtr(params[0]));
|
2003-07-28 01:44:38 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// Write to walkthrough file (zebug0.txt)
|
2008-01-28 00:14:17 +00:00
|
|
|
debug(5, "PLAYING SEQUENCE \"%s\"", filename);
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2007-02-08 21:55:37 +00:00
|
|
|
// don't want to carry on streaming game music when cutscene starts!
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
fnStopMusic(NULL);
|
2003-07-28 01:44:38 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// pause sfx during sequence
|
|
|
|
_vm->_sound->pauseFx();
|
2003-10-18 08:11:50 +00:00
|
|
|
|
2009-02-21 15:07:05 +00:00
|
|
|
_moviePlayer = makeMoviePlayer(filename, _vm, _vm->_mixer, _vm->_system);
|
2003-09-17 13:03:13 +00:00
|
|
|
|
2009-03-07 15:29:52 +00:00
|
|
|
if (_moviePlayer && _moviePlayer->load(filename)) {
|
2008-11-09 13:20:43 +00:00
|
|
|
_moviePlayer->play(_sequenceTextList, _sequenceTextLines, _smackerLeadIn, _smackerLeadOut);
|
2006-07-08 11:42:07 +00:00
|
|
|
}
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2007-02-08 21:55:37 +00:00
|
|
|
_sequenceTextLines = 0;
|
|
|
|
|
2008-11-09 13:20:43 +00:00
|
|
|
delete _moviePlayer;
|
|
|
|
_moviePlayer = NULL;
|
2003-07-28 01:44:38 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// unpause sound fx again, in case we're staying in same location
|
|
|
|
_vm->_sound->unpauseFx();
|
2003-09-17 13:03:13 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
_smackerLeadIn = 0;
|
|
|
|
_smackerLeadOut = 0;
|
2003-07-28 01:44:38 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// now clear the screen in case the Sequence was quitted (using ESC)
|
|
|
|
// rather than fading down to black
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_screen->clearScene();
|
2003-07-28 01:44:38 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// zero the entire palette in case we're about to fade up!
|
2003-07-28 01:44:38 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
byte pal[4 * 256];
|
2003-07-28 01:44:38 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
memset(pal, 0, sizeof(pal));
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_screen->setPalette(0, 256, pal, RDPAL_INSTANT);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
debug(5, "fnPlaySequence FINISHED");
|
2003-09-17 13:03:13 +00:00
|
|
|
return IR_CONT;
|
2003-07-28 01:44:38 +00:00
|
|
|
}
|
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
int32 Logic::fnShadedSprite(int32 *params) {
|
|
|
|
// params: 0 pointer to object's graphic structure
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_router->setSpriteShading(decodePtr(params[0]), SHADED_SPRITE);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
2003-09-17 13:03:13 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
int32 Logic::fnUnshadedSprite(int32 *params) {
|
|
|
|
// params: 0 pointer to object's graphic structure
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
_router->setSpriteShading(decodePtr(params[0]), UNSHADED_SPRITE);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
2003-09-17 13:03:13 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
int32 Logic::fnFadeUp(int32 *params) {
|
|
|
|
// params: none
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_screen->waitForFade();
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
if (_vm->_screen->getFadeStatus() == RDFADE_BLACK)
|
|
|
|
_vm->_screen->fadeUp();
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2003-09-17 13:03:13 +00:00
|
|
|
return IR_CONT;
|
2003-07-28 01:44:38 +00:00
|
|
|
}
|
|
|
|
|
2003-10-18 08:11:50 +00:00
|
|
|
int32 Logic::fnDisplayMsg(int32 *params) {
|
2003-09-17 13:03:13 +00:00
|
|
|
// Display a message to the user on the screen.
|
2003-10-18 08:11:50 +00:00
|
|
|
|
|
|
|
// params: 0 Text number of message to be displayed.
|
2003-09-17 13:03:13 +00:00
|
|
|
|
|
|
|
uint32 local_text = params[0] & 0xffff;
|
|
|
|
uint32 text_res = params[0] / SIZE;
|
2003-07-28 01:44:38 +00:00
|
|
|
|
|
|
|
// Display message for three seconds.
|
2003-09-17 13:03:13 +00:00
|
|
|
|
|
|
|
// +2 to skip the encoded text number in the first 2 chars; 3 is
|
|
|
|
// duration in seconds
|
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_screen->displayMsg(_vm->fetchTextLine(_vm->_resman->openResource(text_res), local_text) + 2, 3);
|
2003-11-16 14:18:29 +00:00
|
|
|
_vm->_resman->closeResource(text_res);
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2003-09-17 13:03:13 +00:00
|
|
|
return IR_CONT;
|
2003-07-28 01:44:38 +00:00
|
|
|
}
|
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
int32 Logic::fnSetObjectHeld(int32 *params) {
|
|
|
|
// params: 0 luggage icon to set
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
uint32 res = (uint32)params[0];
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_mouse->setObjectHeld(res);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnAddSequenceText(int32 *params) {
|
|
|
|
// params: 0 text number
|
|
|
|
// 1 frame number to start the text displaying
|
|
|
|
// 2 frame number to stop the text dispalying
|
|
|
|
|
2007-02-08 21:55:37 +00:00
|
|
|
if (!readVar(DEMO)) {
|
|
|
|
assert(_sequenceTextLines < MAX_SEQUENCE_TEXT_LINES);
|
|
|
|
|
2009-02-21 15:07:05 +00:00
|
|
|
_sequenceTextList[_sequenceTextLines].reset();
|
|
|
|
_sequenceTextList[_sequenceTextLines]._textNumber = params[0];
|
|
|
|
_sequenceTextList[_sequenceTextLines]._startFrame = params[1];
|
|
|
|
_sequenceTextList[_sequenceTextLines]._endFrame = params[2];
|
2007-02-08 21:55:37 +00:00
|
|
|
_sequenceTextLines++;
|
|
|
|
}
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
2003-10-18 08:11:50 +00:00
|
|
|
int32 Logic::fnResetGlobals(int32 *params) {
|
|
|
|
// fnResetGlobals is used by the demo - so it can loop back & restart
|
|
|
|
// itself
|
|
|
|
|
|
|
|
// params: none
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
ScreenInfo *screenInfo = _vm->_screen->getScreenInfo();
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
byte *globals = _vm->_resman->openResource(1) + ResHeader::size();
|
|
|
|
int32 size = _vm->_resman->fetchLen(1) - ResHeader::size();
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2003-10-18 08:11:50 +00:00
|
|
|
debug(5, "globals size: %d", size);
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2003-09-17 13:03:13 +00:00
|
|
|
// blank each global variable
|
2003-10-18 08:11:50 +00:00
|
|
|
memset(globals, 0, size);
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2003-11-16 14:18:29 +00:00
|
|
|
_vm->_resman->closeResource(1);
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2003-09-17 13:03:13 +00:00
|
|
|
// all objects but george
|
2003-11-16 14:18:29 +00:00
|
|
|
_vm->_resman->killAllObjects(false);
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2003-10-07 07:07:47 +00:00
|
|
|
// FOR THE DEMO - FORCE THE SCROLLING TO BE RESET!
|
2003-10-18 08:11:50 +00:00
|
|
|
// - this is taken from fnInitBackground
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2003-09-17 13:03:13 +00:00
|
|
|
// switch on scrolling (2 means first time on screen)
|
2005-02-19 14:02:16 +00:00
|
|
|
screenInfo->scroll_flag = 2;
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2005-10-17 06:45:54 +00:00
|
|
|
// Used to be IR_CONT, but that's a bad idea. We may just have killed
|
|
|
|
// our own script resource -- continuing will cause a bad memory read
|
|
|
|
// access.
|
|
|
|
return IR_STOP;
|
2003-09-17 13:03:13 +00:00
|
|
|
}
|
2003-07-28 01:44:38 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
int32 Logic::fnSetPalette(int32 *params) {
|
|
|
|
// params: 0 resource number of palette file, or 0 if it's to be
|
|
|
|
// the palette from the current screen
|
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_screen->setFullPalette(params[0]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
// use this in the object's service script prior to registering the mouse area
|
|
|
|
// ie. before fnRegisterMouse or fnRegisterFrame
|
|
|
|
// - best if kept at very top of service script
|
|
|
|
|
|
|
|
int32 Logic::fnRegisterPointerText(int32 *params) {
|
|
|
|
// params: 0 local id of text line to use as pointer text
|
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_mouse->registerPointerText(params[0]);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnFetchWait(int32 *params) {
|
|
|
|
// Fetches a resource in the background but prevents the script from
|
|
|
|
// continuing until the resource is in memory.
|
|
|
|
|
|
|
|
// params: 0 resource to fetch [guess]
|
|
|
|
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnRelease(int32 *params) {
|
|
|
|
// Releases a resource from memory. Used for freeing memory for
|
|
|
|
// sprites that have just been used and will not be used again.
|
|
|
|
// Sometimes it is better to kick out a sprite straight away so that
|
|
|
|
// the memory can be used for more frequent animations.
|
|
|
|
|
|
|
|
// params: 0 resource to release [guess]
|
|
|
|
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnPrepareMusic(int32 *params) {
|
|
|
|
// params: 1 id of music to prepare [guess]
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnSoundFetch(int32 *params) {
|
|
|
|
// params: 0 id of sound to fetch [guess]
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnSmackerLeadIn(int32 *params) {
|
|
|
|
// params: 0 id of lead-in music
|
|
|
|
|
|
|
|
// ready for use in fnPlaySequence
|
|
|
|
_smackerLeadIn = params[0];
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnSmackerLeadOut(int32 *params) {
|
|
|
|
// params: 0 id of lead-out music
|
|
|
|
|
|
|
|
// ready for use in fnPlaySequence
|
|
|
|
_smackerLeadOut = params[0];
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Stops all FX and clears the entire FX queue.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnStopAllFx(int32 *params) {
|
|
|
|
// params: none
|
|
|
|
|
2007-01-26 18:21:41 +00:00
|
|
|
_vm->_sound->clearFxQueue(false);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnCheckPlayerActivity(int32 *params) {
|
|
|
|
// Used to decide when to trigger music cues described as "no player
|
|
|
|
// activity for a while"
|
|
|
|
|
|
|
|
// params: 0 threshold delay in seconds, ie. what we want to
|
|
|
|
// check the actual delay against
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
uint32 seconds = (uint32)params[0];
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_mouse->checkPlayerActivity(seconds);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnResetPlayerActivityDelay(int32 *params) {
|
|
|
|
// Use if you want to deliberately reset the "no player activity"
|
|
|
|
// counter for any reason
|
|
|
|
|
|
|
|
// params: none
|
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_mouse->resetPlayerActivityDelay();
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnCheckMusicPlaying(int32 *params) {
|
|
|
|
// params: none
|
|
|
|
|
|
|
|
// sets result to no. of seconds of current tune remaining
|
|
|
|
// or 0 if no music playing
|
|
|
|
|
|
|
|
// in seconds, rounded up to the nearest second
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
writeVar(RESULT, _vm->_sound->musicTimeRemaining());
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
2003-10-18 08:11:50 +00:00
|
|
|
int32 Logic::fnPlayCredits(int32 *params) {
|
2003-09-17 13:03:13 +00:00
|
|
|
// This function just quits the game if this is the playable demo, ie.
|
|
|
|
// credits are NOT played in the demo any more!
|
2003-07-28 01:44:38 +00:00
|
|
|
|
2003-10-18 08:11:50 +00:00
|
|
|
// params: none
|
|
|
|
|
Applied my own patch #1341495, in an attempt to fix alignment issues
reported by Crilith.
To elaborate a bit, the engine no longer accesses resource data through
packed structs. Instead it uses memory streams and the READ/WRITE
functions.
If data is mainly read, not written, I have replaced the old struct with a
new one with a read() function to read the whole thing from memory into the
struct's variables, and a write() function to dump the struct's variables
to memory. In fact, most of these write() functions remain unused.
If data is both read and written, I have replaced the struct with a class
with individual get/set functions to replace the old variables. This
manipulates memory directly.
Since I'm fairly sure that these structs are frequently stored as local
variables for a script, all script variables (both local and global) are
stored as little-endian and accessed through the READ/WRITE functions,
rather than being treated as arrays of 32-bit integers.
On a positive note, the functions for doing endian conversion of resources
and save games have been removed, and some general cleanups have been made
to assist in the rewrite.
Initial reports indicate that this patch indeed fixes alignment issues, and
that I have not - surprisingly - broken the game on big-endian platforms.
At least not in any immediately obvious way. And there's still plenty of
time to fix regressions before 0.9.0, too.
svn-id: r19366
2005-10-29 21:24:54 +00:00
|
|
|
if (readVar(DEMO)) {
|
2008-07-13 22:00:39 +00:00
|
|
|
_vm->quitGame();
|
2004-03-01 00:32:47 +00:00
|
|
|
return IR_STOP;
|
2003-12-31 13:14:07 +00:00
|
|
|
}
|
2003-09-23 06:27:58 +00:00
|
|
|
|
2005-05-02 05:41:01 +00:00
|
|
|
_vm->_screen->rollCredits();
|
2003-09-17 13:03:13 +00:00
|
|
|
return IR_CONT;
|
2003-07-28 01:44:38 +00:00
|
|
|
}
|
2003-10-04 00:52:27 +00:00
|
|
|
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
int32 Logic::fnSetScrollSpeedNormal(int32 *params) {
|
|
|
|
// params: none
|
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_screen->setScrollFraction(16);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnSetScrollSpeedSlow(int32 *params) {
|
|
|
|
// params: none
|
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_screen->setScrollFraction(32);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
// Called from speech scripts to remove the chooser bar when it's not
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// appropriate to keep it displayed
|
|
|
|
|
|
|
|
int32 Logic::fnRemoveChooser(int32 *params) {
|
|
|
|
// params: none
|
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_mouse->hideMenu(RDMENU_BOTTOM);
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Alter the volume and pan of a currently playing FX
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnSetFxVolAndPan(int32 *params) {
|
|
|
|
// params: 0 id of fx (ie. the id returned in 'result' from
|
|
|
|
// fnPlayFx
|
|
|
|
// 1 new volume (0..16)
|
|
|
|
// 2 new pan (-16..16)
|
|
|
|
|
|
|
|
debug(5, "fnSetFxVolAndPan(%d, %d, %d)", params[0], params[1], params[2]);
|
|
|
|
|
|
|
|
_vm->_sound->setFxIdVolumePan(params[0], params[1], params[2]);
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Alter the volume of a currently playing FX
|
|
|
|
*/
|
|
|
|
|
|
|
|
int32 Logic::fnSetFxVol(int32 *params) {
|
|
|
|
// params: 0 id of fx (ie. the id returned in 'result' from
|
|
|
|
// fnPlayFx
|
|
|
|
// 1 new volume (0..16)
|
|
|
|
|
|
|
|
_vm->_sound->setFxIdVolumePan(params[0], params[1]);
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnRestoreGame(int32 *params) {
|
|
|
|
// params: none
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnRefreshInventory(int32 *params) {
|
2005-02-19 14:02:16 +00:00
|
|
|
// Called from 'menu_look_or_combine' script in 'menu_master' object
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
// to update the menu to display a combined object while George runs
|
|
|
|
// voice-over. Note that 'object_held' must be set to the graphic of
|
|
|
|
// the combined object
|
|
|
|
|
|
|
|
// params: none
|
|
|
|
|
2005-02-19 14:02:16 +00:00
|
|
|
_vm->_mouse->refreshInventory();
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
|
|
|
int32 Logic::fnChangeShadows(int32 *params) {
|
|
|
|
// params: none
|
2005-02-19 14:02:16 +00:00
|
|
|
ScreenInfo *screenInfo = _vm->_screen->getScreenInfo();
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
|
|
|
|
// if last screen was using a shading mask (see below)
|
2005-02-19 14:02:16 +00:00
|
|
|
if (screenInfo->mask_flag) {
|
|
|
|
uint32 rv = _vm->_screen->closeLightMask();
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
if (rv)
|
|
|
|
error("Driver Error %.8x", rv);
|
2005-02-19 14:02:16 +00:00
|
|
|
screenInfo->mask_flag = false;
|
Began what I hope is the final major restructuring of the BS2 engine.
In this first step, I have moved all opcode functions into functions.cpp,
instead of having them scattered all over the place.
To get things to compile again, I had to rewrite the overly complicated
sound effects handling. It's much simpler now.
The next step will be to move any non-trivial code out of the opcode
functions and into the appropriate object. This, I hope, will make it
easier to create well-separated objects, instead of the current mess.
I also want to tear down the artificial boundary between the main directory
and the "driver" directory. We already have a cross-platform layer; there's
no need to have yet another one. (Actually, the rewriting of the sound
effects code took one first step in this direction.)
At the final stage, I'd like to get rid of the "drivers" directory
completely, but I'll probably need some help with that if I want to
preserve the CVS history of the code.
Things will probably be a bit bumpy along the way, but I seem to have
reached a point of relative stability again, which is why I'm commiting
this now.
svn-id: r16668
2005-01-28 16:33:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
return IR_CONT;
|
|
|
|
}
|
|
|
|
|
2003-10-04 00:52:27 +00:00
|
|
|
} // End of namespace Sword2
|