This heuristic was originally how all SCI16 speed tests were handled.
It has been gradually replaced with script patches, until all games
were patched in: ea48986006
At the time, I left this in because it had the benefit of speeding up
the SCI11 test variants so that they didn't produce a startup delay.
Now we know that this heuristic has been identifying regular rooms as
speed tests and unthrottling them too, causing unintended effects.
Some of this behavior was masked by fast-cast throttling occurring
everywhere, until: e09010f7d8
For example, the QFG1VGA Sierra logo animation changes speed and runs
very fast as soon as the sparkle is finished. The Longbow map rooms
were also detected as speed test rooms and animated too fast. Cast-less
rooms like LB2's title screen run unthrottled and consume CPU.
There are only a few SCI11 speed test rooms, so now they're explicitly
unthrottled in kGameIsRestarting with the other throttling exceptions.
In 2010 a 10 ms delay was added to every kGetEvent call to prevent
maxing out the CPU when some games display message boxes:
b0b4ddcc52
This rapid kGetEvent polling is called fast cast and it only exists in
SCI1.1 games and some SCI1 games. It occurs in an inner loop when the
fast cast global is set.
Now the workaround is only applied to fast cast games and only when the
kGetEvent polling occurs.
Removing this 10 ms delay from every SCI16 game cycle doesn't change
game speed by itself because the larger dynamic throttling in
kGameIsRestarting has been absorbing the 10 ms. But this does make the
other throttles easier to understand and work with, and solves minor
edge cases where delays from multiple kGetEvent calls add up.
Disables SCI16 speed tests with generic script patches.
- All games now pass their speed tests, and consistently.
- Handles games/versions that the benchmarking heuristic misses.
- Fixes bug #13529 (speed test overflow on fast CPUs) on all games.
- Replaces most game-specific patches.
I've left the benchmarking heuristic so that any game or version that
still needs patching will still be handled with unthrottling.
The new approach is to delay kAnimate briefly (with an 68ms timeout)
while there is a zero-tick PalVary running, so that it has time to
trigger.
The previous workaround would immediately process a zero-tick
PalVaryInit/PalVaryReverse. This caused problems in QfG3 (bug #10304)
where it interfered with PalVaryPause.
The previous workaround could also be modified to handle pause/resume,
but this new approach should be closer to SSCI's behaviour, which used a
timer for a zero-tick PalVary too.
This fixes bug #10304, and keeps #5298 fixed too.
- Add "kAnimate fast cast state" to "version" debug command
- Make it possible for script patcher signatures to get fully
used outside of the regular script patcher
- Remove previous fastcast detections and replace them with
a signature heuristic
- Remove object name checking, when fastcast global is set
- Heuristic detects "fast cast" support incorrectly for multilingual
KQ5, but it seems the game never sets the global, so it won't
matter. KQ5 CD (also SCI1 late) has fastcast support.
- Remove hack in GfxView::draw
- Add lots of comments to ScriptPatcher class
This fixes EcoQuest 1 Floppy showing the anemone on top of the
message box (see bug #5170)
No functionality change has been made with this commit. This avoids
setting and getting the reg_t members directly, and is the basis of any
future work on large SCI3 scripts (larger than 64KB)
This tries to make our code a bit more compliant with our code formatting
conventions. For future use, this is the command I used:
git ls-files "*.cpp" "*.h" | xargs sed -i -e 's/[ \t]*$//'
This renaming allows us to better distinguish that this version is for games
that only had an EGA version, and avoid confusion with newer SCI1 game releases
with EGA graphics (e.g. KQ5 EGA). The only game with this SCI version is QFG2,
a SCI1 EGA game with a parser. Also, added some games for each SCI version.
adjusted comments about that and also renamed the bit, also added additional TODO, because hoyle 4 has special code that gets into action, when bit 0 is not set, but bit 2 is
svn-id: r53853
kAddToPic doesnt do loop/cel fixups, it also doesnt support global scaling and will only set nsRect on scaled Cels, not reusing fill() of kAnimate anymore, but doing those things separately
fixes kq6 - 2nd window shown instead of peppermint bush (bug #3070465)
should be save to backport
svn-id: r52887
it's set. Currently game-specific for Hoyle4; fixes bug #3038424. But
other SCI1.1 games should be tested with a general version of this to
see whether it fixes relevant bugs.
svn-id: r52818
lb2floppy: scripts get confused in the rat scene, when scaleX/Y are set with applied global scaling. This results in kInitBresen to get called almost endlessly, which results in ego slowing down to a crawl and not being able to reach the upper right exit - fixes bug #3050606
svn-id: r52559
SCI constantly invoked Engine::shouldQuit(), which in
turn called 2 virtual functions. This added a significant
overhead, as this was called constantly without any
throttling whatsoever. Now, the invocation of shouldQuit()
is throttled to be on each frame update (i.e. at a rate of
60fps). Thanks to wjp for profiling this.
svn-id: r52431
fixes kq5 trunk getting closed again during intro - is actually a script bug that sets cel to some really high number - bug #3037578, may also fix other similar issues
svn-id: r51969
sluicebox
Eugene Sandulenko
Orgad Shaneh
Kawa
Willem Jan Palenstijn
Colin Snover
Martin Kiewitz
Alexandre Detiste
Johannes Schickel
Filippos Karapetis
Filippos Karapetis
Max Horn
Matthew Hoops
strangerke
Lars Skovlund