Fixes Trac#10575.
When stepping away or pressing the buttom for the mechanical age fortress
rotator after setting it west it would be south. This is because some variables
used for a workaround were being reset everytime the rotator puzzle
was being reinitalized (in o_fortressRotation_int). Which happens when
the player steps back up to the rotator controls.
This change simply removes the variables being reset so they retain
the variables they had before. Those variables are already initalized
in the Mechanical constructor.
The direction that the fortress is pointing is calculated when
the user locks on using the left throttle (after the gears have
stopped). It also happens when the user hits the button.
There is only a difference in the two behaviors when the user has
changed the direction, but not locked on.
The 2nd case can lead to the direction be different than
what the last locked on direction was. Since the user
can lock the direction to say north, turn it to west, not
lock on and hit the button, which will set the direction
to west. This can lead to confusion since a direction sound
isn't played in this case.
This change removes code, to prevent the recalculation
of the direction when the button is pressed so that direction
is always what it was last locked onto.
The observatory go button computes random 105x106 rectangles in the
512x512 star image. The previous code would allow to go out of bounds by
3 pixels. Interestingly the bug is in the original engine disassembly
perhaps it could crash as well.
Fixes Trac#10576.
Fixes Trac#10572.
The day slider does not go all the way to the bottom (day 31st)
when the down arrow is used to increment the day. This change
increases how much y displacement happens with each change
in the day so that the bottom is reach on day 31st.
This is just a graphical bug and doesn't affect the logic
of the puzzle. The day slider can also be dragged to the bottom
without using the buttons.
In D'ni, after giving the white page to Atrus, or after coming without a
page, too videos of Atrus would play on top of each other after waiting
long enough for the first video to loop.
Fixes#10503.
The global variable "ending" tracks various
states that Atrus can be in.
This variable determines what video of
Atrus will be playing and if the brother
books have been destroyed.
Despite this variable being called ending
there are other endings not captured by
this variable (like the brother endings).
Also change spelling of Dini to Dni for one
usage of ActiveAge from a previous commit.
This global basically tracks what age/book/stack
the player is in.
It is also used to change stacks via
MohawkEngine_Myst::changeToStack() by
being an index for MystScriptParser::_stackMap
and MystScriptParser::_startCard.
E.x., instead of _global.currentAge = 2
now _global.currentAge = kMystLibrary.
Also two of the endings are considered Ages
instead of being apart of _global.ending.
In the original the sound reciever would provide a hint of the
source sound direction when the user was less than 5 degrees
away. In ScummVM it was less than or equal to 5 degrees.
Now it matches the behavior of the original.
This slows down the Selentic age sound receiver spin rate
when fully depressing either of the turn buttons.
This is a deviation from the original and improves
the users experience.
In the original when the button is fully depressed
the receiver turns so fast that it is impossible
to predict what angle you full be at when you let go.
Fixes #Trac10486.
For 12 Hr format versions, in ScummVM when the hour was set to
what should be 12:XX it was displayed as 0:XX.
This differs with the original English versions which displays
the 12:xx.
I verified this fix in both Myst ME and Myst English CD.
Change all inner loops to call the same frame function.
Remove the draw delay code. Now that we have proper frame pacing it is
not necessary anymore to wait before drawing to make sure the previous
frame stayed on the screen long enough.
Fixes#9915, #9922, #9923.
