November 2006 - Wine-devel - list.winehq.org

Wine 0.9.25 full of regressions!
by Mirek 21 Nov '06

21 Nov '06

Hi, i just instaled wine 0.9.25 (i had 0.9.24) and it is full of regressions, why is this possible, why is not there someone who can test wine before it came out to new release? 1. 3DMark 2003 - cant open setigns menu 2. 3DMark 2003 - with GLSL, almost all working test are broken (bad graphic) 3. 3DMark 2003 - with ARB is about 25% slower 4. HL2 Episode One - Some characters are black 5. NFS: MW - with ARB it is slower 6. W3 - cannot run, because of Opengl init 7. W3 FT - same as W3 8. nVidiaSDK - opengl demos in nVidia are not working anymore 9. nVidiaSDK - some Direct3D demos are broken more than in release 0.9.24 9. Prey - not working This is pretty big list, almost half of my apps are not working, or working worst than before, it is sad. Mirek Slugen

10 19

Re: winequartz.drv: Add mouse events handling and stub window creation support. (take2)
by Detlef Riekenberg 21 Nov '06

21 Nov '06

On Mo, 2006-11-20 at 16:48 +0100, Pierre d'Herbemont wrote: > +/* --------------------------------------------------------------------- > + * send_mouse_input > + */ > +static void send_mouse_input( HWND hwnd, DWORD flags, DWORD x, DWORD y, > + DWORD data, DWORD time, DWORD extra_info, UINT injected_flags ) > +{ > + POINT pt; > + > + pt.x = x; > + pt.y = y; Hi. Some more questions. POINT pt is never used for more than accessing pt.x and pt.y in this function, which is really useless, as both values are fn-parameters. Is there code waiting, that use pt? A comment would be nice for this case. Did you already try to replace "DWORD x, DWORD y" with "POINT pt" for the driver-internal Functions? > +BOOL QDRV_CreateWindow( HWND hwnd, CREATESTRUCTA *cs, BOOL unicode ) > + if (cs->cx > 65535) > + { > + ERR( "invalid window width %d\n", cs->cx ); > + cs->cx = 50; > + } Setting the minimal or maximal size to 50 looks strange for me. Is this an leftover from debugging? > \ No newline at end of file This is easy to fix. -- By by ... Detlef

6 6

Re: rpcrt4: Allow applications to specify authentication levels other than connect in RpcBindingSetAuthInfoA/W.
by Kai Blin 21 Nov '06

21 Nov '06

On Tuesday 21 November 2006 14:09, Robert Shearman wrote: > + /* FIXME: the mapping should probably be retrieved using SSPI somehow */ > + if (AuthnLevel == RPC_C_AUTHN_LEVEL_DEFAULT) > + AuthnLevel = RPC_C_AUTHN_LEVEL_NONE; I figure this could be done using QueryContextAttributes, I guess. So far this is a bit stubbed out, but it should basically do the right thing (tm) for integrity and confidentiality. Cheers, Kai -- Kai Blin, <kai Dot blin At gmail Dot com> WorldForge developer http://www.worldforge.org/ Wine developer http://wiki.winehq.org/KaiBlin/ -- Will code for cotton.

1 0

Re: user32: Add a rudimentary WM_SETFONT message test for the static control, make it pass under Wine
by Alexandre Julliard 21 Nov '06

21 Nov '06

Dmitry Timoshkov <dmitry(a)codeweavers.com> writes: > @@ -496,7 +504,10 @@ static LRESULT StaticWndProc_common( HWN > { > SetWindowLongPtrW( hwnd, HFONT_GWL_OFFSET, wParam ); > if (LOWORD(lParam)) > - STATIC_TryPaintFcn( hwnd, full_style ); > + { > + InvalidateRect( hwnd, NULL, TRUE ); > + UpdateWindow( hwnd ); > + } A single RedrawWindow call would be more efficient. > @@ -673,9 +684,7 @@ static void STATIC_PaintTextfn( HWND hwn > if (!IsWindowEnabled(hwnd)) SetTextColor(hdc, GetSysColor(COLOR_GRAYTEXT)); > } > > - if (!(len = SendMessageW( hwnd, WM_GETTEXTLENGTH, 0, 0 ))) return; > - if (!(text = HeapAlloc( GetProcessHeap(), 0, (len + 1) * sizeof(WCHAR) ))) return; > - SendMessageW( hwnd, WM_GETTEXT, len + 1, (LPARAM)text ); > + if (!(len = InternalGetWindowText( hwnd, text, sizeof(text)/sizeof(WCHAR) ))) return; You are introducing a hard-coded limit to the length of the text, that's not a good idea. -- Alexandre Julliard julliard(a)winehq.org

1 0

FW: Re: Hebrew translation of wine
by Scott Ritchie 21 Nov '06

21 Nov '06

Of course, translations are always welcome. The trick is, I really have no idea how Rosetta works or which parts belong in Ubuntu Rosetta and which parts belong in upstream Wine. Therefore I'm forwarding your message to wine-devel, hopefully someone there can enlighten me. Thanks, Scott Ritchie On Mon, 2006-11-20 at 18:53 +0200, Amichai Teumim wrote: > Hi Scott > > I would like to translate Wine to Hebrew for Ubuntu using Rosetta. > > Would you approve of this? > > Thanks > > Amichai

4 4

Re: d3d [9]: Remove AddRef from IWineD3DResource_GetDevice.
by Markus Amsler 21 Nov '06

21 Nov '06

Don't apply d3d[9] - d3d[14], they are not independent and I missed one GetDevice AddRef. But I strongly assume they were already dropped for other reasons. Anything wrong with them? Markus

1 0

re: OLE Automatization Linux's OpenOffice from wine app
by Dan Kegel 21 Nov '06

21 Nov '06

Kochmin wrote: > I have windows application wich use OLE Automatization to fill-in > OpenOffice document It's create OleObject like > OpenOffice := CreateOleObject('com.sun.star.ServiceManager'); >.. > How can I rewrite it, or use any tools, to work when OpenOffice is > installed on Linux, and my application runs in wine? Try installing the Windows version of OpenOffice in Wine. - Dan

4 5

re: What operating systems will Winelib run on non-x86 systems?
by Dan Kegel 21 Nov '06

21 Nov '06

Hi Wei Li, last year, you asked in http://www.winehq.org/hypermail/wine-users/2004/06/0086.html > I could not find the info regarding what operating systems > (such as AIX 5L v5.2, HP-UX 11i, Solaris Release Level > 9) will Winelib run on non-x86 machines. I'll really > appreciate it if someone can provide my related link or > info. I can only find the following link regarding the > supporting OS but it does not provide Winelib's supporting > OS info: > http://www.winehq.org/site/docs/wine-faq/index 3.1. > Under what hardware > platform(s) and operating system(s) will Wine(Lib) run? I'm sorry nobody got back to you -- the wine-devel mailing list might have been a better place to ask. I believe winelib runs on the following non-x86 operating systems: Linux, Solaris, Mac OS X, and FreeBSD. There's no reason in principle it couldn't run on AIX or HP/UX. Some work has been done on an HP/UX port, but some assembly would be required to complete it: http://www.winehq.com/site/?issue=241#HP-UX%20Port Cheers, Dan -- Wine for Windows ISVs: http://kegel.com/wine/isv

3 2

OpenAL and DirectSound
by Stefan Dösinger 21 Nov '06

21 Nov '06

Hi, Just stumbled uppon some interesting thing about openal and DirectSound 3D on vista: http://www.openal.org/openal_vista.html It sounds like openal will be the old way to get hw accelerated 3d sound on vista, and creative is writing a dsound -> openal wrapper. Maybe it gets a proper license so we can include it in wine :-) Maybe we should write an openal thunk for wine? Many games use openal, and it seems that on wine openal uses directsound. That might work more better and faster if we had a openal.dll which directly called linux openal :-) Just dropping the suggestions :-) Stefan

3 3

Regression in wininet between 0.9.23 and 0.9.24
by Pavel Troller 21 Nov '06

21 Nov '06

Hi! I would like to report a regression in wininet.dll, causing Dreamcom Kubik SMS to crash as follows: wine: Unhandled page fault on read access to 0x00000000 at address 0x7e687b86 (thread 0013), starting debugger... Unhandled exception: page fault on read access to 0x00000000 in 32-bit code (0x7e687b86). Register dump: CS:0023 SS:002b DS:002b ES:002b FS:0063 GS:006b EIP:7e687b86 ESP:7d936820 EBP:7d9368d8 EFLAGS:00010246( - 00 -RIZP1) EAX:00000016 EBX:7e6a6728 ECX:00000002 EDX:00000000 ESI:00000000 EDI:00000000 Stack dump: 0x7d936820: 7b9177e0 7d936834 00000000 7e6a2520 0x7d936830: 001f3538 7e6a6728 001f3538 20000013 0x7d936840: 7d936900 7e68763b 00000000 f7cc0a2e 0x7d936850: 00000013 00000000 01160230 00000000 0x7d936860: 00010000 7b87fac6 7d9368b8 7bc35945 0x7d936870: 00001257 00010000 0010bdd0 7e68e354 Backtrace: =>1 0x7e687b86 in wininet (+0x17b86) (0x7d9368d8) 2 0x7e687f6b HttpQueryInfoW+0x1cb in wininet (0x7d936918) 3 0x7e6884d0 HttpQueryInfoA+0xc0 in wininet (0x7d936968) 4 0x0057a62c in dreamcom (+0x17a62c) (0x7d9369c8) 5 0x00579fb0 in dreamcom (+0x179fb0) (0x7d9369f0) 6 0x00578d0b in dreamcom (+0x178d0b) (0x7d9369f8) 7 0x7b88b488 in kernel32 (+0x6b488) (0x7d936ac8) 8 0x7bc5b570 in ntdll (+0x4b570) (0x7d9373c8) fixme:dbghelp:elf_load_debug_info_from_map Alpha-support for Dwarf2 information for libpthread.so.0 9 0xf7dd5528 start_thread+0xb8 in libpthread.so.0 (0x7d9374b8) 10 0xf7d6021e __clone+0x5e in libc.so.6 (0x00000000) 0x7e687b86: cmpw $0,0x0(%edx) I was doing "cvs bisect" in dlls/wininet and found that the problem lies between .23 and .24 - from .24 on it crashes, .23 is still OK. So now I'm using HEAD wine with .23 wininet and the program is happy. With regards, Pavel Troller

2 1