April 2011 - Wine-Bugs - list.winehq.org

[Bug 3930] HOMM shows black screen (sound problem)
by wine-bugs＠winehq.org April 26, 2011

April 26, 2011

http://bugs.winehq.org/show_bug.cgi?id=3930 --- Comment #46 from Michał Z. Gołębiowski <m.goleb+winehq.org(a)gmail.com> 2011-04-26 09:58:56 CDT --- I need a little help here. I can't patch Wine 1.3 as it has a weird bug - HoMM1 screen displays only a upper-left quarter of the proper graphics scaled to full screen. I wanted to apply it to Wine 1.2. Of course I had to modify a diff a bit so let me provide the patch, it was applied to the version 1.2.2-1ubuntu1~maverick2 from ubuntu-wine PPA. Now when I try to run wine, I get this: $ WINEPREFIX="$HOME/.wine-homm1" /home/mgol/Programs/wine1.2-source/wine1.2-1.2.2/wine Heroes.exe /I0 wine: cannot find L"C:\\windows\\system32\\wineboot.exe" err:process:start_wineboot failed to start wineboot, err 2 err:module:import_dll Library WINMM.dll (which is needed by L"C:\\Games\\HoMM1\\Heroes.exe") not found err:module:import_dll Library ws2_32.dll (which is needed by L"c:\\windows\\system32\\netapi32.dll") not found err:module:import_dll Library NETAPI32.dll (which is needed by L"C:\\Games\\HoMM1\\Heroes.exe") not found err:module:import_dll Library WINMM.dll (which is needed by L"C:\\Games\\HoMM1\\wail32.dll") not found err:module:import_dll Library wail32.dll (which is needed by L"C:\\Games\\HoMM1\\Heroes.exe") not found err:module:import_dll Library WINMM.dll (which is needed by L"C:\\Games\\HoMM1\\wail32.dll") not found err:module:import_dll Library wail32.dll (which is needed by L"C:\\Games\\HoMM1\\smkwai32.dll") not found err:module:import_dll Library WINMM.dll (which is needed by L"C:\\Games\\HoMM1\\smkwai32.dll") not found err:module:import_dll Library smkwai32.dll (which is needed by L"C:\\Games\\HoMM1\\Heroes.exe") not found err:module:LdrInitializeThunk Main exe initialization for L"C:\\Games\\HoMM1\\Heroes.exe" failed, status c0000135 though e.g. wineboot.exe exists on a given path. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 0

[Bug 8051] Sims 2 demo exits prematurely
by wine-bugs＠winehq.org April 26, 2011

April 26, 2011

http://bugs.winehq.org/show_bug.cgi?id=8051 --- Comment #112 from Henri Verbeet <hverbeet(a)gmail.com> 2011-04-26 07:34:49 CDT --- (In reply to comment #111) > Wrt Microsoft Office & friends: Yeah, there are monetary incentives for that. > Lots of CodeWeavers customers use that app. Wrt WoW, the fixes are likely to > benefit other apps. Monetary incentives could help Sims 2 as well, but the > amount we're talking about would probably in the tens of thousands of dollars > given the amount of work needed for SW vertex processing. That's a CodeWeavers limitation of being sensitive to that kind of consideration. It doesn't necessarily hold for Wine in general. In the general case you just need to find someone who's willing to spend a couple of months implementing this, either because that person really likes Sims 2 or because code generation is interesting, and actually has a chance of getting the work committed. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 0

[Bug 8051] Sims 2 demo exits prematurely
by wine-bugs＠winehq.org April 26, 2011

April 26, 2011

http://bugs.winehq.org/show_bug.cgi?id=8051 --- Comment #111 from Stefan Dösinger <stefandoesinger(a)gmx.at> 2011-04-26 07:21:41 CDT --- We don't need SW emulation for fragment shaders. Windows doesn't offer that, at least not in a way that it can be used for "productive" playing(there's the refrast, but it is not a replacement for GPU fragment processing). But still you want a fairly sophisticated solution for SW shaders that includes generating x86 / x86_64 code. An interpreter will probably not give satisfying performance, especially on older hardware where SW vp is intended to replace missing GPU functionality. So LLVM sounds like a good idea. Integrating LLVM into Wine is already a challenge. The other challenge is integrating SW shaders into the existing wined3d codebase in a clear way. The SW shader stuff should be a new shader backend, and you'll have to make sure the shader input(Steam sources) are read properly. The shader output(processed vertices) has to be passed to the opengl fixed function pipeline. The gl pipeline has to be set up in a way that the processed vertex data is passed through to the fragment pipeline properly, similarly to how RHW vertices are treated now. Wrt Microsoft Office & friends: Yeah, there are monetary incentives for that. Lots of CodeWeavers customers use that app. Wrt WoW, the fixes are likely to benefit other apps. Monetary incentives could help Sims 2 as well, but the amount we're talking about would probably in the tens of thousands of dollars given the amount of work needed for SW vertex processing. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 0

[Bug 8051] Sims 2 demo exits prematurely
by wine-bugs＠winehq.org April 26, 2011

April 26, 2011

http://bugs.winehq.org/show_bug.cgi?id=8051 --- Comment #110 from Henri Verbeet <hverbeet(a)gmail.com> 2011-04-26 07:03:42 CDT --- Well, the way to fix this is to implement support for software shaders. That shouldn't be terribly hard, but still at least a couple of months of work for vertex shaders, and probably more than that for fragment shaders. You'd probably want to use LLVM. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 0

[Bug 8051] Sims 2 demo exits prematurely
by wine-bugs＠winehq.org April 26, 2011

April 26, 2011

http://bugs.winehq.org/show_bug.cgi?id=8051 Abraham <stormbolter(a)gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |stormbolter(a)gmail.com --- Comment #109 from Abraham <stormbolter(a)gmail.com> 2011-04-26 06:57:47 CDT --- (In reply to comment #108) > (In reply to comment #107) > > Last I checked, Wine hardcodes the constant instead of actually checking the > > hardware. > Where would that be? Wined3d reads the constants from the opengl library. It > subtracts a number of constants from the available ones on certain drivers(e.g. > fglrx and OSX) where we know that the driver needs some for its own purposes. > This doesn't change the number of constants that is advertised to the d3d app > though. The driver reserved constants are there to avoid GLSL compile errors > with shaders that use indirect addressing(and thus potentially use all > available constants) > > We do limit the hardware shader constants to 256 in d3d9 and d3d8 because there > are some games that break if more constants are supported. > d3d9:Set*ShaderConstantF rejects constants >= 256 as well. This seems incorrect > with software vertex processing and we should probably change the check a bit. > (e.g. check against the HW capabilities, and if hardware VP is used cap the > constants at 256, otherwise 8192) > > > But what I've never understood about this bug is, lots of bugs are > > filed that prevent an application or game from starting, and they might be the > > only affected app/game, but people pay attention and usually attempt to fix the > > problem. This bug for some reason has been demoted its entire life as > > unimportant. > There are three things in play here: (a) the effort needed for a proper fix, > (b) the limited effect on other games(we'd probably just fix a single game) and > (c) that implementing a proper fix requires a lot of knowledge how wined3d > works. > > (a)+(b) prevent the main wined3d developers from working on this bug, (c) makes > it unlikely that a new volunteer shows up and writes a patch. I think (B) is not very relevant. There are a lot of fixes made to run a single piece of software (ex: WOW or Microsoft Word), that probably affects only the single app or just a simple subset. I know I'm just voicing my frustration here. There is little save putting a small reward that I can do to help this bug fix, but is the single bug that refrains me from migrating my girlfriend's computer to ubuntu, as she is a very avid sims 2 gamer (and doesn't like sims 3 -too many micropayments, and too much money invested on previous game-). I can replace office, and teach her to use shotwell to manage her photo collection, but if she has to reboot to play the sims 2, is no use. Many sims 2 players are like my girlfriend, avid fans, but no tech skills (I'm no programmer myself). A suggestion... There's any mechanism for offering bounties for bugfixing? Maybe, while we cannot offer tech expertise, we can offer a monetary compensation... -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 0

[Bug 8051] Sims 2 demo exits prematurely
by wine-bugs＠winehq.org April 26, 2011

April 26, 2011

http://bugs.winehq.org/show_bug.cgi?id=8051 --- Comment #108 from Stefan Dösinger <stefandoesinger(a)gmx.at> 2011-04-26 05:28:42 CDT --- (In reply to comment #107) > Last I checked, Wine hardcodes the constant instead of actually checking the > hardware. Where would that be? Wined3d reads the constants from the opengl library. It subtracts a number of constants from the available ones on certain drivers(e.g. fglrx and OSX) where we know that the driver needs some for its own purposes. This doesn't change the number of constants that is advertised to the d3d app though. The driver reserved constants are there to avoid GLSL compile errors with shaders that use indirect addressing(and thus potentially use all available constants) We do limit the hardware shader constants to 256 in d3d9 and d3d8 because there are some games that break if more constants are supported. d3d9:Set*ShaderConstantF rejects constants >= 256 as well. This seems incorrect with software vertex processing and we should probably change the check a bit. (e.g. check against the HW capabilities, and if hardware VP is used cap the constants at 256, otherwise 8192) > But what I've never understood about this bug is, lots of bugs are > filed that prevent an application or game from starting, and they might be the > only affected app/game, but people pay attention and usually attempt to fix the > problem. This bug for some reason has been demoted its entire life as > unimportant. There are three things in play here: (a) the effort needed for a proper fix, (b) the limited effect on other games(we'd probably just fix a single game) and (c) that implementing a proper fix requires a lot of knowledge how wined3d works. (a)+(b) prevent the main wined3d developers from working on this bug, (c) makes it unlikely that a new volunteer shows up and writes a patch. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 0

[Bug 3930] HOMM shows black screen (sound problem)
by wine-bugs＠winehq.org April 26, 2011

April 26, 2011

http://bugs.winehq.org/show_bug.cgi?id=3930 --- Comment #45 from Damjan Jovanovic <damjan.jov(a)gmail.com> 2011-04-26 02:23:39 CDT --- (In reply to comment #44) > Are those patches going to be integrated? I don't really like to compile wine > by myself, then I'd have to repeat the process each time upgraded package pops > up. (especially that I have some weird problems with wine1.3, so I prefer to > use wine1.2) MCI is complex system, with 16 and 32 bit interfaces and drivers, dating back to at least Windows 3.1. The only way any patch is going to be integrated is if it does the right thing, and the patches here are just hacks that work around the problem. An investigation into what exactly Windows does is necessary, and nobody has done it yet: you'd probably have to test each Windows version individually. It is worth noting that Heroes of Might and Magic doesn't work on eg. Windows 2000. In the meanwhile, Heroes of Might and Magic is a long game with a high replay value. IMHO it is well worth the time you put into patching and compiling a separate Wine prefix just for it. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 0

[Bug 25792] New: Warhammer 40k Dark Crusade single player campaign issue
by wine-bugs＠winehq.org April 26, 2011

April 26, 2011

http://bugs.winehq.org/show_bug.cgi?id=25792 Summary: Warhammer 40k Dark Crusade single player campaign issue Product: Wine Version: 1.3.11 Platform: x86 OS/Version: Linux Status: UNCONFIRMED Severity: blocker Priority: P2 Component: -unknown AssignedTo: wine-bugs(a)winehq.org ReportedBy: alexandrbezenkov(a)gmail.com Game crashes right after the global map of the single player campaign is showed. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 14

[Bug 26929] New: Self compiled wine is crashing planescape torment, with stacktrace
by wine-bugs＠winehq.org April 26, 2011

April 26, 2011

http://bugs.winehq.org/show_bug.cgi?id=26929 Summary: Self compiled wine is crashing planescape torment, with stacktrace Product: Wine Version: 1.3.18 Platform: x86-64 OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown AssignedTo: wine-bugs(a)winehq.org ReportedBy: i30817(a)gmail.com Hello, as the title says, a self compile is crashing planescape torment. I believe this happens on intro video, and i thought it might be useful for the "crash on rest" bug (that is still reproducible on ppa supplied wine). I opened a new bug since this is happening always, whether i install the fixpack or not, so it appears another bug. Hopefully it is not just a compile error, and it is useful. wine head compiled with "CC="ccache gcc -m32" LDFLFLAGS="-L./lib32" CFLAGS="-g -O0" ./configure --verbose --disable-tests && make depend -j 3 && make -j 3" Removed ~/.wine. wineconfig settings different from default are virtual desktop 800x600 and Directsound hardware acceleration set to Emulation. Here is the log of the exe ~/wine-git/wine Torment.exe > bug.txt 2>&1 -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 7

[Bug 8051] Sims 2 demo exits prematurely
by wine-bugs＠winehq.org April 25, 2011

April 25, 2011

http://bugs.winehq.org/show_bug.cgi?id=8051 --- Comment #107 from Robert Showalter <rshowalter0520(a)verizon.net> 2011-04-25 18:50:02 CDT --- (In reply to comment #105) > * it is not clear if the graphic bugs are caused by the hack or is entirely > unrelated. The hack is not in wine because it is an incorrect hack, not because > Sims 2 doesn't render properly with it. Even if Sims2 worked perfectly with > that hack it would not be committed. Yup. Who implied that this hack should be committed? It was purely to demonstrate what causes the game to fail to start. > * DirectX 10 capable nvidia cards(geforce 8+) support 1024 vertex shader > constants, so they should work around the SW vertex processing bug. This does > not apply to AMD dx10 cards. Last I checked, Wine hardcodes the constant instead of actually checking the hardware. But what I've never understood about this bug is, lots of bugs are filed that prevent an application or game from starting, and they might be the only affected app/game, but people pay attention and usually attempt to fix the problem. This bug for some reason has been demoted its entire life as unimportant. Is it because of the complexity of it or just because it just doesn't seem as important as a missing reference in a DLL? I would think any known deficiency, whether a popular one or not, should be fixed simply to keep Wine's performance as close to its targeted imitation as possible. I don't mean this as an attack or anything, and I appreciate the HUGE undertaking of Wine as a whole, I'm just trying to understand the politics. Can't say that Sims 2 is too old and out of date... a lot of people still play Sims 2 over Sims 3, if for no other reason than the massive amount of custom content available to it, with still more being made. Although I will admit, I don't think this bug alone will solve all the problems. Last time I tried to implement the hack in Wine, there were reports of some operation on stencils not being supported. This may have to do with the visibility issues. The final glaring problem of Sims being spread eagle and glowing I can't even begin to guess... although a similar thing happen when a texture fails to load on an object, although that's a glowing blue effect. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 0