July 2009 - Wine-Devel - list.winehq.org

MaxVertexBlendMatrices issue
by Karl Vogel July 11, 2009

July 11, 2009

Just tracked down a crash in the game Dark Age of Camelot (daoc) when pixelshaders are enabled. The game issues a GetDeviceCaps() and then checks if MaxVertexBlendMatrices >= 4, which on wine is set to zero on my videocard (Geforce 6800), whereas a windows box returns 4. I found this old post concerning MaxVertexBlendMatrices causing issues in another game (EQ): http://www.winehq.org/pipermail/wine-devel/2008-February/062432.html http://www.winehq.org/pipermail/wine-devel/2008-February/062437.html So I was wondering what the proper way to fix this would be? Should this be made overrideable by a registry setting for certain games? Or should the GL check be changed? Or is that software emulated vertex blending patch needed to correctly fix the issue? Bugzilla has 3 other games where the simple 'fix' of returning 4 makes the game work (Warhammer, EQ, Shin Megami Tensei Online): http://bugs.winehq.org/buglist.cgi?quicksearch=MaxVertexBlendMatrices -- For reference the daoc bug is http://bugs.winehq.org/show_bug.cgi?id=10832

1 0

GSoC mid-term evaluations due
by Kai Blin July 11, 2009

July 11, 2009

Hi GSoC students and mentors, Just a little reminder that you need to fill out your GSoC surveys by monday. Cheers, Kai -- Kai Blin WorldForge developer http://www.worldforge.org/ Wine developer http://wiki.winehq.org/KaiBlin Samba team member http://www.samba.org/samba/team/ -- Will code for cotton.

1 0

[2/2] msxml3: Register Missing Components
by Lady Bug July 11, 2009

July 11, 2009

2 1

Re: [1/2](resend)advapi: break the core of LookupAccountNameW out into helper functions.
by Juan Lang July 10, 2009

July 10, 2009

Hi Aric, it strikes me that this could be broken up a bit further. This'd make it rather easier to review. For example, --- a/dlls/advapi32/advapi32_misc.h +++ b/dlls/advapi32/advapi32_misc.h @@ -24,4 +24,15 @@ const char * debugstr_sid(PSID sid); BOOL ADVAPI_IsLocalComputer(LPCWSTR ServerName); BOOL ADVAPI_GetComputerSid(PSID sid); +BOOL lookup_local_wellknown_name(LPCWSTR lpAccountName, DWORD ccAccountName, + PSID Sid, LPDWORD cbSid, + LPWSTR ReferencedDomainName, + LPDWORD cchReferencedDomainName, + PSID_NAME_USE peUse, BOOL *ret); +BOOL lookup_local_user_name(LPCWSTR lpAccountName, DWORD ccAccountName, + PSID Sid, LPDWORD cbSid, + LPWSTR ReferencedDomainName, + LPDWORD cchReferencedDomainName, + PSID_NAME_USE peUse, BOOL *ret); In the initial patch, these aren't referenced outside of security.c, so leaving them static there would make the changes smaller. A subsequent patch that does reference them from lsa.c could make them non-static. +static inline WCHAR *strnrchrW(const WCHAR *str, WCHAR ch, DWORD *len) It's not clear to me what this function does from its name, and you didn't document it. There's no corresponding strnrchr libc function, so that doesn't help either. It appears from reading it that you made a function like strrchrW that returns the location of the found ch in str, like strrchrW does, but also sets *len to the position at which it's found? If that's correct, you could accomplish the same thing by calling strrchrW and doing pointer arithmetic with its return value. These suggestions notwithstanding, it seems like introducing this helper function is one change, and deserving of its own patch. + +/* + * Helper function for LookupAccountNameW and LookupNames2 in lsa.c The second part of that comment, that it's a helper for LookupNames2, isn't correct in this first patch. It becomes true later, but the comment should be changed then, not in the first patch. (Minor nit: mentioning the file LookupNames2 is in seems superfluous.) + * returns TRUE if the AccountName is handled (even if with an error). + * returns FALSE if not. + */ + +BOOL lookup_local_wellknown_name(LPCWSTR lpAccountName, DWORD ccAccountName, + PSID Sid, LPDWORD cbSid, + LPWSTR ReferencedDomainName, + LPDWORD cchReferencedDomainName, + PSID_NAME_USE peUse, BOOL *ret) The use of two BOOLs is confusing to me. For one thing, it makes the patch larger: - ret = CreateWellKnownSid(ACCOUNT_SIDS[i].type, NULL, pSid, &sidLen); + *ret = CreateWellKnownSid(ACCOUNT_SIDS[i].type, NULL, pSid, &sidLen); This change could be avoided if you returned the old meaning of ret, and used the BOOL * parameter to indicate whether the AccountName is handled. Finally, I'd suggest that breaking out looking up well known names could be separate from breaking out looking up local names. Hope that helps, --Juan

1 0

Re: winemine: Add ellipses for menu entries that open new dialog windows (DE)
by Paul Vriens July 10, 2009

July 10, 2009

Detlef Riekenberg wrote: > There are more winemine translations without the ellipses. > Can the changes all in one Patch? > > There are non UTF8 - resources with special characters. > Should they change to UTF8 first? > > > > > ------------------------------------------------------------------------ > > Hi Detlef, There is patch in the queue that actually gets rid of some ellipses as there was a discussion about this on wine-devel. -- Cheers, Paul.

1 0

Warnings on the translation page
by Paul Vriens July 10, 2009

July 10, 2009

Hi, I found an anomaly on the translations page. Whenever the Makefile shows a generic rc file (last one added for RC_SRCS) the language before that shows warnings on the page. Example: http://source.winehq.org/transl/lang.php?lang=004%3A01 zh (Chinese (Traditional)) shows up most of the warning as the entries in the Makefile are sorted. Not sure yet, if this is an issue with wrc or transl. -- Cheers, Paul.

2 4

Wine very very slow in Chinese language of Linux
by axel_xia＠sohu.com July 10, 2009

July 10, 2009

Dear Madam / Sir, How are you? I am writing email to you for the issue probably caused by Chinese fonts. When using WINE in Chinese environment Linux, the WINE works very very slow even opening windows / clicking buttons.... I search the problem in google, found lots of users have same issue regardless Fedora or Ubuntu or other distributed Linux OS. Someone presented the resolution by changing Chinese fonts configuration because they think WINE is always searching suitable GKB font sets (which sometimes are not exisiting in Linux OS), someone said it was related to X11 (not compatiable or some linux OS not using X11)..., and others guess the issue related to other reason..... So, dear Madam / Sir, could you please help the WINE fans who are using Linux in Chinese environment to figure out this kind of issue? Or is that possible to put the fixed WINE into Yum or someother software database? Hopefully, the user of different distributed Linux can be of no issue when WINE M$ software in Chinese? Thanks and best regards! Yours sincerely, Axel

1 0

Re: wined3d: Fix some checkGLcalls.
by Henri Verbeet July 10, 2009

July 10, 2009

Most of those looked good enough to me to begin with, but I guess it doesn't hurt.

1 0

Re: Check if WINED3DRS_VERTEXBLEND is dirty in transform_world
by Henri Verbeet July 9, 2009

July 9, 2009

2009/7/5 Lisa Denia <eiffel56(a)gmail.com>: > This fixes the flickering cursor mentioned in Bug > #14375(http://bugs.winehq.org/show_bug.cgi?id=14375). > The cursor flickers because the scene gets rendered twice a draw, which is > known and expected. The cursor is not one of that usual animated 2d things, > but a 3d object. So, in every draw, the scene gets rendered, the cursor gets > rendered, and the scene gets rendered again. The attached patch checks for > that special case. > It's not entirely clear to me why this should work.

2 1

Re: kernel32/nls: Add Lithuanian winerr translation
by Paul Vriens July 9, 2009

July 9, 2009

Hi Aurimas, Congratulations !! -- Cheers, Paul.

2 1

image loading with windowscodecs
by Vincent Povirk July 9, 2009

July 9, 2009

So I've been working on an implementation of windowscodecs.dll (Windows Imaging Component). This is a Vista/.NET 3.x API for manipulating image data. To the best of my knowledge, only 2 programs currently use it, and I'm not running either one. We're not likely to see it used by real programs in Wine much, since there's no redistributable. A program using it would have to require Vista or .NET 3.0, or require users to download it manually. It's documented here: http://msdn.microsoft.com/en-us/library/ms735422(VS.85).aspx Adding this dll will not increase compatibility, at least not directly. It will, however, provide Wine with a single component that can be used by any dll or program that needs to load or save a raster image. All of the code for png, gif, jpeg, etc should be migrated into this dll. If we need any image formats not supported by native, we can implement them as Wine-specific extensions. We need to do this because oleaut32 (the current home of image-loading code) sucks. I have been told that the libpng code should not be duplicated throughout the codebase. I need features in gdiplus that oleaut32 does not provide, such as access to the image dpi and metadata, and the ability to save images. In fact, much of the gdiplus image/bitmap api matches WIC 1:1. On Windows, gdiplus is known to be able to use WIC extensions to support new image types, so in theory it should use WIC (though this probably won't matter any time soon). I'm told that image loading is also useful for d3dx9 dll's. The WIC api is well-designed in my opinion and should be able to provide any operations needed by other components to load/save images. The next step is to implement the image types needed by oleaut32 and start using WIC to load them in olepicture, starting with BMP. To that end, I need a good test program. I need to see what kind of BMP files currently work (I suspect that many are broken) and what kinds will work if I replace the BMP-loading code. This will tell me at what point I can make the change without introducing a regression. After that, I'll need to do the same thing with the other file types. So I need a general-purpose image viewer that uses either gdiplus or oleaut32 to do its image loading. Does anyone know any? -- Vincent Povirk

1 0

Re: [1/2] winemine: Add ellipses for menu entries that open new dialog windows
by Yuri Khan July 9, 2009

July 9, 2009

On Thu, Jul 9, 2009 at 15:53, Anders Jonsson<anders.jonsson(a)norsjonet.se> wrote: > POPUP "&Info" { > - MENUITEM "&Fastest Times", IDM_TIMES > + MENUITEM "&Fastest Times...", IDM_TIMES > MENUITEM "&About", IDM_ABOUT > } According to Microsoft's guidelines, an ellipsis is only necessary when the command asks for confirmation or additional parameters. "If simply viewing the window is a common action, don't use an ellipsis." http://msdn.microsoft.com/en-us/library/aa511502.aspx#ellipses

2 1

Re: transl and sublangs are not showing neutral
by Ricardo Filipe July 9, 2009

July 9, 2009

2009/7/8 Paul Vriens <paul.vriens.wine(a)gmail.com> > Paul Vriens wrote: > >> Ricardo Filipe wrote: >> >>> hi mikolaj, >>> >>> in transl the Portuguese language (not Portugese :( ) is not showing most >>> of the neutral strings for it's sublangs (portugal and brazil). >>> i think this is not how it should be... in each sublang it should show >>> the Neutral derived strings and the warning that neutral is being used and >>> should not be used aside the language name in the main list, not put neutral >>> separate in another hidden page. >>> >>> as it is now it's hard for me to know what is translated and what's not. >>> what do you (and wine devel) think? could you please change this? >>> >>> regards, >>> Ricardo >>> >>> >>> ------------------------------------------------------------------------ >>> >>> >>> Hi Ricardo, >> >> So you would be happy/happier if we show "Portuguese (Neutral)" (with the >> warning) on the main page next to the "Portuguese (Portugal)" and >> "Portuguese (Brazilian)"? >> >> Just re-read your email. What you like is to have "Portuguese (Portugal)" > show the combination of "Portuguese (Portugal)" and "Portuguese (Neutral)", > correct? With a warning for the neutral resources. > > What does this mean for "English (Neutral)" should that be treated the same > or is Portuguese an exception? > > -- > Cheers, > > Paul. > yes that is what i would like and think is correct showing. portuguese strings should defer to neutral if none is found in the sublang, with a warning that they may be incorrect. i would say english is the exception by having all translated in SUBLANG_DEFAULT and not separating sublangs as is done with portuguese resources and others. as i stated in irc one of these days, i think the solution to english is to have a SUBLANG_BRITISH or whatever and not assume neutral is for british english, how it is right now contradicts the use of wrc which is sublang -> neutral -> default (english). i'd say english should have all resources in neutral and then the different ones in sublang_british or sublang_australian or other. at least that is what makes sense to me for the process to be consistent.

2 6

Re: [PATCH 1/7] jscript: Inherit Function_valueOf from Object
by Piotr Caban July 9, 2009

July 9, 2009

Hi, Please drop this bunch of patches (it will not apply to newest git). Piotr

1 0

Re: WineD3D: Check if GL accepts 4 component specular colors
by Henri Verbeet July 9, 2009

July 9, 2009

Actually, while I'm at it, you're also missing ENTER_GL()/LEAVE_GL() calls in match_allows_spec_alpha(), and you should document that it depends on the caller to do context activation.

1 0

Re: comctl32: Fix dialog ('Next' button was hidden)
by Aurimas Fišeras July 9, 2009

July 9, 2009

On 07/08/2009 10:31 PM, Frédéric Delanoy wrote: > I think it should be under the button "Finish", so that the "Finish" button can be shown in the same place instead of "Next >" on the last wizard page.

3 3

Re: WineD3D: Check if GL accepts 4 component specular colors
by Henri Verbeet July 9, 2009

July 9, 2009

2009/7/9 Stefan Dösinger <stefan(a)codeweavers.com>: > - GL_EXTCALL(glSecondaryColorPointerEXT)(e->format_desc->gl_vtx_format, e->format_desc->gl_vtx_type, > - e->stride, e->data + stateblock->loadBaseVertexIndex * e->stride + offset[e->stream_idx]); > + > + if(GLINFO_LOCATION.quirks & WINED3D_QUIRK_ALLOWS_SPECULAR_ALPHA) > + { > + /* The spec disallows 4 component specular colors, but most GL drivers accept it. Try to use it, > + * since the 4th value can be passed through to pixel shaders > + */ > + GL_EXTCALL(glSecondaryColorPointerEXT)(e->format_desc->gl_vtx_format, e->format_desc->gl_vtx_type, > + e->stride, e->data + stateblock->loadBaseVertexIndex * e->stride + offset[e->stream_idx]); > + } > + else > + { > + GLenum type = e->format_desc->gl_vtx_type; > + GLint format = e->format_desc->gl_vtx_format; > + if(format == 4 && type == GL_UNSIGNED_BYTE) > + { > + GL_EXTCALL(glSecondaryColorPointerEXT)(3, GL_UNSIGNED_BYTE, > + e->stride, e->data + stateblock->loadBaseVertexIndex * e->stride + offset[e->stream_idx]); > + } > + else > + { > + FIXME("Add specular color code for type %x, fmt %u\n", type, format); > + GL_EXTCALL(glSecondaryColorPointerEXT)(format, type, > + e->stride, e->data + stateblock->loadBaseVertexIndex * e->stride + offset[e->stream_idx]); > + } > + } I think you should test for the specific case the quirk addresses. For example, 3 component specular color will trigger the FIXME with this code. Also, EXT_vertex_array_bgra adds GL_BGRA as a valid type/size parameter for glSecondaryColorPointer(). Just test for "type == 4" and handle it according to the quirk being present or not. Combinations that are always invalid, like e.g. 1 or 2 component data should already have been filtered by declaration_element_valid_ffp().

1 0

Re: comctl32: Fix dialog ('Next' button was hidden) [ invalid patch ]
by Frédéric Delanoy July 9, 2009

July 9, 2009

Please discard this patch Frédéric

1 0

Translations/locale
by Ken Sharp July 9, 2009

July 9, 2009

Can someone tell me what's going on on this page http://source.winehq.org/transl/lang.php?lang=009%3A00 ? If you click on the bottom links (locales) I see a message "Invalid resource file". Does this mean the resource file doesn't exist, or is there a little oops in the links? Thanks, Ken.

2 2

Re: winex11.drv: Handle clipboard on an auxiliary thread for windowless apps (try 3)
by Dmitry Timoshkov July 9, 2009

July 9, 2009

"Yuri Khan" <yurivkhan(a)gmail.com> wrote: > +void selection_acquire(void) Should be 'static'. > +DWORD WINAPI selection_thread_proc(LPVOID lphwnd) Should be 'static'. What's the purpose of such a parameter name? 'dummy', 'unused' or just 'param' would be better IMO. > +{ > + HANDLE handles[1]; > + > + selection_acquire(); > + > + while (selectionAcquired) > + { > + MsgWaitForMultipleObjectsEx(0, handles, INFINITE, QS_SENDMESSAGE, 0); > + } What's the purpose of 'handles'? 'handles' is not initialized, moreover the number of passed handles is 0. MsgWaitForMultipleObjectsEx() works just fine with '0, NULL' combination. -- Dmitry.

1 0

Fwd: [Wine] Tmax Window(a propietary OS) using Wine?
by Austin English July 8, 2009

July 8, 2009

---------- Forwarded message ---------- From: nengad24 <wineforum-user(a)winehq.org> Date: Tue, Jul 7, 2009 at 5:54 AM Subject: [Wine] Tmax Window(a propietary OS) using Wine? To: wine-users(a)winehq.org Hi, Im posting here the next doubt: Yesterday a korean company has released a new OS which is compatible with MS Windows Apps. As they state, they are a propietary OS, and i cant see any GPL code or SVN repository.Of course my lack of Korean can help to not find the open source code,but the little im able to read seems they arent going to release any source code. I know that is quite difficult to replicate Windows APIs, so i´m asking myself if this OS is using Wine code to replicate the APIs. If they are using, they could be violating GPL rules. http://www.tmaxwindow.co.kr/ If you want to have more Info, i point you to the ReactOS Forum,where a thread about this Company can be found. http://www.reactos.org/forum/viewtopic.php?f=2&t=7134&st=0&sk=t&sd=a TMax wanted to support MS Windows drivers "using ReactOS as reference(sic)" thanks to "ReactOS is GPL code" but seems they werent able to "replicate" and that they just use a Kernel(i didnt find which one)and "their Code" to support Windows Apps. I know there is somekind of not good relation between Wine and ReactOS, but it´s my obligation to aware you about this issue since we (ReactOS+Wine) are members of the Open Source Community. Wishing that my feels are wrong,and that any GPL has been violated, One Open Source Community Guy. -- -Austin

5 6

Font problem in Notepad -- regression today
by Susan Cragin July 8, 2009

July 8, 2009

Notepad is behaving very strangely today. It crashes every time I try to change the font. However, when I tried to trace the problem using winedbg, the font changed fine, after the following fixme showed up. Then when I closed the winedbg terminal window, notepad closed too. susan(a)ubuntu:~$ winedbg Wine-dbg>info process pid threads parent executable (all id:s are in hex) 00000008 1 00000000 'notepad.exe' 0000000e 3 0000000a 'services.exe' 00000011 4 0000000e 'winedevice.exe' 00000018 1 00000008 'explorer.exe' Wine-dbg>attach 0x8 0x680007f0 GLIBC_2+0x7f0 in ld-linux.so.2: int $0x80 Wine-dbg>set $BreakOnFirstChance=0 Wine-dbg>cont fixme:winedbg:be_i386_is_func_call Unsupported yet call insn (0xcd) at 0x680007f0

1 0

Obsolete print dialogs in comdlg32?
by Frédéric Delanoy July 8, 2009

July 8, 2009

Hi, During update of translation of comdlg32 dialogs, I saw that some printing dialogs seem broken, namely PRINT and PRINT_SETUP, while PRINT32 and PRINT32_SETUP look OK. I checked some programs and the PRINT[_SETUP] dialogs don't seem to be used (a casual check in the code didn't find usages). Are they still needed? Or can they simply be deleted? Frédéric

2 1

Re: winhttp(2/5): Simplify netconn_resolve when using getaddrinfo
by Hans Leidekker July 8, 2009

July 8, 2009

Hi Juan, - *sa_len = sizeof(struct sockaddr_in); - memset( sa, 0, sizeof(struct sockaddr_in) ); - memcpy( &((struct sockaddr_in *)sa)->sin_addr, &((struct sockaddr_in *)res->ai_addr)->sin_addr, sizeof(struct in_addr) ); - ((struct sockaddr_in *)sa)->sin_family = res->ai_family; - ((struct sockaddr_in *)sa)->sin_port = htons( port ); + *sa_len = res->ai_addrlen; + memcpy( sa, res->ai_addr, res->ai_addrlen ); This causes test failures, you can't ignore the port. -Hans

2 1

Re: [comctl32/tests] Fix test failures with comctl32 <= 5.80
by Nikolay Sivov July 8, 2009

July 8, 2009

Paul Vriens wrote: > Hi, > > Most of them are with comctl32 <= 5.80 but there seem to be some > different failures for 4.70. And not all failures for <= 5.80 are on > 4.70. I didn't think changing the comment for that is a big deal. > > I don't have 4.70 on any of my boxes but I've tested with 4.72. > > Changelog > Fix test failures with comctl32 <= 5.80 > You forgot a patch.

1 0