| Age | Commit message (Collapse) | Author |
|---|
|
They pursued it with forks and hope;
They threatened its life with a railway-share;
They charmed it with smiles and soap."
Maybe this time the elusive cursor hotspot bugs will finally be fixed, once
and for all. The blitCursor() function now calculates both the real and the
virtual dimensions of the cursor, since they may both be different from the
original dimensions.
svn-id: r22913
|
|
svn-id: r22849
|
|
scaling or aspect-ratio correction, so it has to be added using real surface
coordinates. So I had to re-introduce the extra parameter to addDirtyRect().
svn-id: r22681
|
|
svn-id: r22679
|
|
svn-id: r22633
|
|
svn-id: r22631
|
|
each iteration of the loop, causing a serious slowdown on my computer. It
empathically should not be doing that, of course, but to safe-guard against
such patological cases, if the entire screen is dirtied even once we can set
_forceFull to true.
svn-id: r22616
|
|
svn-id: r22585
|
|
svn-id: r22562
|
|
svn-id: r22551
|
|
svn-id: r22550
|
|
svn-id: r22542
|
|
think _cursorHasOwnPalette is needed any more, so I've removed it.
svn-id: r22541
|
|
cursor palette in overlay mode. Now it does, which fixes bug #1476666. The
reason it worked at all before was that blitCursor() was called while still not
in overlay mode.
This, of course, means that the GUI can corrupt the in-game cursor palette. I
don't have any game that uses this feature, so I don't know if it's properly
restored again afterwards. But if it isn't, that should be fairly easy to fix.
svn-id: r22539
|
|
svn-id: r22538
|
|
svn-id: r22535
|
|
- Disable scaling for cursors which have target scale defined
svn-id: r22509
|
|
- Eriktorbjorn's patch from same tracker item for scaling sword1/2 cutscenes
is applied as is. It lacks resolution switch on-the-fly.
- GUI widgets are repositioned on the fly and use most space, even aspect
ratio corrected screen is used without scaling
- Heavy tesing is required, but works for me in all cases except for bug
#1483272: "GUI: SCUMM pause dialog breaks upon scaler switch" which needs more
work.
- I probavly broke some backend or two
svn-id: r22505
|
|
- Renamed base/options.cpp to base/commandLine.cpp because of conflict
with gui/options.cpp which sit in same directory in MSVC builds
- Moved AudioCDManager singleton declaration outside of Audio namespace
- Fixed numerous MSVC warning of potentially uninitialized variables and
int <-> bool conversions.
svn-id: r22397
|
|
svn-id: r22071
|
|
- Icon (icons/scummvm.info) contains stack requirement so the stack cookie
can be removed
svn-id: r21863
|
|
icons/ and all other assorted stuff in dists/ directory. Still I may broke some
packaging/building procedures as no testing was performed.
svn-id: r21859
|
|
svn-id: r21710
|
|
svn-id: r21605
|
|
moved the call to quit() from scummvm_main to the various backend main routines (porters may want to replace it by something different)
svn-id: r21559
|
|
OSystem class before calling scummvm_main (Note: PalmOS and Symbian are not yet converted, and won't work currently)
svn-id: r21557
|
|
Remove NO_CONSOLE, since it is never defined
svn-id: r21554
|
|
svn-id: r21553
|
|
he might remove it completly in a future version)
svn-id: r21550
|
|
base/main.cpp to sdl.cpp, where it belongs (considering that it is backend specific code)
svn-id: r21547
|
|
various backend specific hacks unnecessary. As a consequence, it is now the
responsibility of the backend to define main. Hence I adapted the SDL backend
accordingly.
svn-id: r21542
|
|
svn-id: r21385
|
|
the reason(s) SDL_SetVideoMode failed
svn-id: r21291
|
|
svn-id: r21055
|
|
the Lure popup menus won't work.
svn-id: r20915
|
|
svn-id: r20904
|
|
svn-id: r20797
|
|
svn-id: r20786
|
|
and x11 backends; added engines/module.mk
svn-id: r20584
|
|
svn-id: r20535
|
|
svn-id: r20424
|
|
svn-id: r20088
|
|
attempts to use invalid rects for blitting
svn-id: r19521
|
|
loadGfxMode(). As reported on the forum, setting aspect ratio correction
in a 640x480 game would cause ScummVM to crash when the game is started
from the launcher.
With all the graphics transactions and stuff, it's a bit hard to follow
what's going to happen, but I hope deferring this check until the last
moment (i.e. just before the graphics surfaces are actually created) is the
correct bugfix.
svn-id: r19491
|
|
--disable-sdl-parachute should be completely undocumented and not mentioned
to end-users. Wrong usage may lead to system lock-up especially when used
with fullscreen mode.
SDL Parachute is used to prevent system instability and do a graceful exit
if something bad happens. It catches most of signals sent to thr process.
Side effect of that that quite often when SDL application dies, core dump
isn't made on *nix systems. This feature adds possibility to disable it.
svn-id: r19146
|
|
svn-id: r19142
|
|
line endings are correct since I have to use TortoiseCVS for this update :)
svn-id: r19115
|
|
except essential Normal1x. Also thumbnails get compiled unconditionally
so we get them on all platforms, even those which never display them.
Sorted out mess brought to sdl backend by epoc maintainer. He misused
__SYMBIAN32__ where DISABLE_SCALERS is more appropriate.
svn-id: r18850
|
|
svn-id: r18689
|
|
svn-id: r18604
|
