#6502 RINGWORLD2: Graphics glitches when going to Ringworld

[Torbjörn Andersson]

English CD version of the game

In the "Meanwhile..." screen that appears when going to Ringworld, there are what looks to me like some slight graphics glitches near the bottom of the screen. I have not compared this to the original, however.

There's also what looks like a palette glitch when it fades down, but I wasn't able to catch a screenshot of that.

[Torbjörn Andersson]

Ignore the white line at the bottom. That's probably part of the window decoration, left over when I cropped the image.

[Torbjörn Andersson]

I have verified in DOSBox that these glitches do not appear in the original. Another thing that's missing in ScummVM is that in the cutscene that follows, whenever the (mostly) unseen ship fires, the screen is supposed to flash white.

[Strangerke]

Could you provide a savegame just before that cutscene? Thanks in advance!

[Torbjörn Andersson]

There is one attached to the first comment I made. Simply give the infodisc to seeker, then speak to him.

[digitall]

Replicated with the attached savegame and my copy of R2R on Linux x86_64 with the latest git master.

Running under Valgrind, it also occurs. There are two valgrind warnings which may be relevant:
==3364== Conditional jump or move depends on uninitialised value(s)
==3364== at 0x4E49E34: SDL_RLESurface (in /usr/lib64/libSDL-1.2.so.0.11.4)
==3364== by 0x6DE9F3: SurfaceSdlGraphicsManager::blitCursor() (surfacesdl-gra
phics.cpp:1947)
==3364== by 0x6DD21F: SurfaceSdlGraphicsManager::setPalette(unsigned char con
st, unsigned int, unsigned int) (surfacesdl-graphics.cpp:1456)
==3364== by 0x42861D: TsAGE::ScenePalette::fade(unsigned char const, bool, int) (core.cpp:1466)
==3364== by 0x5175F3: TsAGE::Ringworld2::Scene1000::signal() (ringworld2_scenes1.cpp:123)
==3364== by 0x423E09: TsAGE::Action::remove() (core.cpp:206)
==3364== by 0x65D11E: TsAGE::SequenceManager::remove() (converse.cpp:88)
==3364== by 0x65D1E1: TsAGE::SequenceManager::signal() (converse.cpp:100)
==3364== by 0x423F0D: TsAGE::Action::dispatch() (core.cpp:226)
==3364== by 0x5B346D: TsAGE::Scene::dispatch() (scenes.cpp:328)
==3364== by 0x518D0C: TsAGE::Ringworld2::Scene1000::dispatch() (ringworld2_scenes1.cpp:472)
==3364== by 0x4342CE: TsAGE::SceneHandler::dispatch() (core.cpp:4459)
==3364==

==3364== Conditional jump or move depends on uninitialised value(s)
==3364== at 0x4E49E34: SDL_RLESurface (in /usr/lib64/libSDL-1.2.so.0.11.4)
==3364== by 0x6DE9F3: SurfaceSdlGraphicsManager::blitCursor() (surfacesdl-gra
phics.cpp:1947)
==3364== by 0x6D961F: SurfaceSdlGraphicsManager::setFeatureState(OSystem::Fea
ture, bool) (surfacesdl-graphics.cpp:234)
==3364== by 0x6CD93D: ModularBackend::setFeatureState(OSystem::Feature, bool)
(modular-backend.cpp:56)
==3364== by 0x6F7E03: Graphics::CursorManager::popCursorPalette() (cursorman.
cpp:190)
==3364== by 0x6F2458: Engine::~Engine() (engine.cpp:155)
==3364== by 0x41F1A2: TsAGE::TSageEngine::~TSageEngine() (tsage.cpp:60)
==3364== by 0x41F1D5: TsAGE::TSageEngine::~TSageEngine() (tsage.cpp:64)
==3364== by 0x40B0B0: runGame(PluginSubclass<metaengine> const*, OSystem&, Co
mmon::String const&) (main.cpp:251)
==3364== by 0x40C03F: scummvm_main (main.cpp:487)
==3364== by 0x409BBF: main (posix-main.cpp:45)
==3364== </metaengine>

Though it looks to me like those artifacts may be leftovers from the starfield in the "Quantum Drive Jump" video that precedes the Meanwhile screen. This could either be due to some issue in Dirty Rects or it could be something related to the video decoder used for the movie?

[digitall]

Yes... looking again. The artifacts look very much like two "stars" from the previous video.