The code was compied from AdvancedDetector. Currently it is
set up to detect Mac indy3/loom which contains it data in
Rooms 1, Roooms 2 and Rooms 3 directories.
svn-id: r49798
FR #2821513: "GUI: add tooltips".
Added tooltips for Add Game button, clear field buttons and couple
other.
Current problem: Only first call correctly restores text. I could
not find where restore information gets lost.
svn-id: r49774
Now it is possible to specify hotkeys in place for ButtonWidget
and CheckboxWidget (the only widgets with hotkeys now).
Use de-facto standard with putting hotkey between tildes (~).
Like '~O~pen'. The tildes gets stripped before after hotkey
is extracted.
This is done for giving translators possibility to specify
their own hotkeys.
Old hotkeys defined at widget instance creation are left for
leaving possibility to specify non-printable hotkeys such as
Common::ASCII_ESCAPE.
Translation files were updated respectively.
svn-id: r49766
Bug #1328120: "MANIACNES: Screen width incorrect, camera halts sometimes".
Fixed by workaround and was tested with intro and on the kitchen,
where now it is possible to see Edna without her noticing you
first.
svn-id: r49764
Bug 1879604: "MANIACNES: Music not played when loading game".
Actually previous attempt to fix this bug was not successful
since first entry of script 5 does not have needed functionality.
Now we resume music manually.
svn-id: r49763
Bug #1438631: "SCUMM: Detecting mac version of indy3/loom broken"
fixed by implementing recursive directory lookup similar to what
was done for AdvancedDetector, since SCUMM engine does not use it.
svn-id: r49656
These are incorrect as '/' operator has precedence over >>, but this does improve readability anyway.
This bug in cppcheck has already been corrected: http://sourceforge.net/apps/trac/cppcheck/ticket/1714
svn-id: r49517
These functions are only used internally be Engine subclasses, and
by moving them to a separate header we can reduce indirect header
dependencies.
svn-id: r48934
The idea behind this is exactly the same as behind r48620,
but it affects all engines, thus engine authors can now use
CursorMan.replaceCursor without having to worry about possible
memory leaks or the like.
svn-id: r48626
The problem here was that SCUMM only uses CursorMan.replaceCursor. When no
cursor had been setup before it would cause the SCUMM cursor to be never
removed from CursorMan, since in this case replaceCursor just uses pushCursor.
To avoid this problem I am just pushing a dummy cursor (and palette, since
that is used in SCUMM HE games too) on engine setup and removing it on engine
destruction.
Actually every engine should setup their first cursor via CursorMan.pushCursor
and then on quit remove it again via CursorMan.popCursor. Using
CursorMan.replaceCursor is *no* good idea for the first cursor to setup, since
that might either replace an existing cursor, thus destroying the caller's
cursor, or pushing a new cursor on the stack, which might result in a leak.
This would also result in making a call to CursorMan.popCursor unsafe, since
that might also impact the caller's cursor setup again.
svn-id: r48620
