- Wine-Devel - list.winehq.org

Warcraft 3 : success and woes
by Jean-Baptiste Note July 20, 2002

July 20, 2002

Hello dear list, ** All the following has been teste on debian sid, XF4.0.2-17, latest wine CVS, on a Geforce2 with 2960 nvidia drivers. GL include were the standart mesa ones. system is pIII-smp, stock 2.4.18 kernel. ** I've followed with great interest a thread on wine-devel where Matthieu Foillard explained his problems trying to run warcraft3 under standard wine. I came across the same problem, so i though i could post here to explain what to do. This is really simple, and following Jesse Allen's recommendation found on http://www.winehq.com/hypermail/wine-devel/2002/07/0189.html The problem appears when you try to run warcraft 3 in opengl mode, regardless of whether or not you ran it with d3d and wineX previously. Warcraft is (following Jesse Allen's thought) trying to scan video modes available in OpenGL and gets stuck. However, after only one successful launch, the game creates the registry entries needed, and does not need to go through this detection cycle again. Question is, how to create those entries ? At your option * try to merge/use my user.reg attached below NOTE: i've still to look to the cd-key entry in my brother's win system to add it and be able to connect to battle.net =) You'll get into trouble certainly though, so read on. * try to do a minimal patch to do one successful run (warning, not very beautiful, we're not on wine-devel): You have to edit the function WINPOS_MinMaximize in the file wine/dlls/x11drv/winpos.c modify it so that it'll return SWP_NOSIZE | SWP_NOMOVE; anyway. compile & install, run. I dont know how this morbid modification affects other things. AFTER the successfull run, you can happily launch the game: launch it and relaunch it change resolution at your leasure BUT... If you do only one small change on window focusing... then it seems impossible for the game window to get the focus again on relaunch. All goes to the launching window ; i've tried with various window managers (kwin, pwm, blackbox) : same problem. So I thought : some information is written somewhere, that is not valid any more when i do a focus change. 1) could it be in user.reg/wine itself ? Deleting ~/.wine/user.reg will solve the problem. However, i've diffed two user.reg differently generated ; i'm at a loss to find any significant difference between the two. (the diff's attached) Furthermore, storing away user.reg and rm -Rf'ing ~/.wine, and copying back, with some focus change in between, leads to a non-focused game window. 2) could it be elsewhere ? IT IS ! As a matter of fact, building a fresh user.reg, then replacing it with an old saved one, without focus changes works flawlessly. I'm pretty sure something somewhere gets written with something like a window number in it. I'm pretty sure that's either in X itself or in the window manager. BTW, the results are the same wether i use wine in managed or unmanaged mode : the WM seems to be unaware of the created window. I don't know if I made myself clear ; i'd be glad to help you (*cough*) fix the bug for me, or (*cough* *cough*) give me some information/pointer as to how i could fix it. Regards, Jean-Baptiste Note

2 1

A windows clone without wine?
by Uwe Bonnes July 20, 2002

July 20, 2002

Hallo, http://slashdot.org/articles/02/07/19/2116234.shtml?tid=99 talks about China developping a windows clone. I can't imagine they don't use wine as a base. Does anybody know more? Bye -- Uwe Bonnes bon(a)elektron.ikp.physik.tu-darmstadt.de Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt --------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------

1 0

Re: PATCH: Convert HTASK to a void*
by Michael Stefaniuc July 19, 2002

July 19, 2002

Hello, please disregard this patch, it has build problems in a "make clean" enviroment. An other problem is that WOWHandle32 and WOWHandle16 are needed in loader/task.c, but I suspect I'm not allowed to import wow32 into the ntdll? bye michael On Fri, Jul 19, 2002 at 01:07:16AM +0200, Michael Stefaniuc wrote: > this patch should close bug #502, but I have some questions (that's why > wine-devel is CC'ed): > - is it correct to use the HTASK_{16,32} macros or should I use > WOWHandle{16,32}? The difference is that the macros include a type > cast to the right handle. > - GetAppCompatFlags() works internaly with a HTASK16. Can I move the > implementation to GetAppCompatFlags16()? Would save a handle > conversion in GetAppCompatFlags16(). > > I'll close the bug report after the patch get's accepted. > > License: LGPL, X11 > Changelog: > Michael Stefaniuc <mstefani(a)redhat.com> > - Convert HTASK to a void* > - Fix HTASK_32 macro > > > bye > michael -- Michael Stefaniuc Tel.: +49-711-96437-199 System Administration Fax.: +49-711-96437-111 Red Hat GmbH Email: mstefani(a)redhat.com Hauptstaetterstr. 58 http://www.redhat.de/ D-70178 Stuttgart

1 1

Wine CVS does *not* compile anymore. [wine/tools/widl//header.c]
by Boris Reisig July 19, 2002

July 19, 2002

Wine stopped compiling a fews days ago or maybe a week. I download the CVS every few days and I noticed that it cant compile anymore. Here make[1]: Entering directory `/root/wine/tools' make[2]: Entering directory `/root/wine/tools/wpp' make[2]: Nothing to be done for `all'. make[2]: Leaving directory `/root/wine/tools/wpp' make[2]: Entering directory `/root/wine/tools/widl' gcc -c -I. -I. -I../../include -I../../include -I../../tools/wpp -g -O2 -Wall -mpreferred-stack-boundary=2 -D__WINE__ -D_REENTRANT -I/usr/X11R6/include -o header.o header.c In file included from header.c:36: ../../tools/wpp/y.tab.h:10: parse error before `wrc_sll_t' ../../tools/wpp/y.tab.h:10: warning: no semicolon at end of struct or union ../../tools/wpp/y.tab.h:11: warning: type defaults to `int' in declaration of `ull' ../../tools/wpp/y.tab.h:11: warning: data definition has no type or storage class ../../tools/wpp/y.tab.h:14: parse error before `cval' ../../tools/wpp/y.tab.h:14: warning: type defaults to `int' in declaration of `cval' ../../tools/wpp/y.tab.h:14: warning: data definition has no type or storage class ../../tools/wpp/y.tab.h:15: parse error before `*' ../../tools/wpp/y.tab.h:15: warning: type defaults to `int' in declaration of `marg' ../../tools/wpp/y.tab.h:15: warning: data definition has no type or storage class ../../tools/wpp/y.tab.h:16: parse error before `*' ../../tools/wpp/y.tab.h:16: warning: type defaults to `int' in declaration of `mtext' ../../tools/wpp/y.tab.h:16: warning: data definition has no type or storage class ../../tools/wpp/y.tab.h:17: parse error before `}' ../../tools/wpp/y.tab.h:17: warning: type defaults to `int' in declaration of `yystype' ../../tools/wpp/y.tab.h:17: warning: data definition has no type or storage class ../../tools/wpp/y.tab.h:66: parse error before `pplval' ../../tools/wpp/y.tab.h:66: warning: type defaults to `int' in declaration of `pplval' ../../tools/wpp/y.tab.h:66: warning: data definition has no type or storage class header.c: In function `is_object': header.c:180: `tOBJECT' undeclared (first use in this function) header.c:180: (Each undeclared identifier is reported only once header.c:180: for each function it appears in.) header.c: In function `is_local': header.c:189: `tLOCAL' undeclared (first use in this function) header.c: In function `is_callas': header.c:198: `tCALLAS' undeclared (first use in this function) make[2]: *** [header.o] Error 1 make[2]: Leaving directory `/root/wine/tools/widl' make[1]: *** [widl] Error 2 make[1]: Leaving directory `/root/wine/tools'

1 0

Problem with wcmd
by Steven Edwards July 19, 2002

July 19, 2002

This patch http://cvs.winehq.com/patch.py?id=1025909276607367637873725 results in this under mingw with gcc 2.95.c and 3.1 gcc -c -I. -I. -I../../include -I../../include -D__MINGW__ -D_WINDOWS -DWINE_NO WINSOCK -Wall -mpreferred-stack-boundary=2 -DSTRICT -DNONAMELESSUNION -DNONAMELE SSSTRUCT -D_REENTRANT -o directory.o directory.c directory.c: In function `WCMD_filesize64': directory.c:369: `uint' undeclared (first use in this function) directory.c:369: (Each undeclared identifier is reported only once directory.c:369: for each function it appears in.) directory.c:369: parse error before `r' directory.c:374: `i' undeclared (first use in this function) directory.c:378: `r' undeclared (first use in this function) make: *** [directory.o] Error 1 Any ideas? Thanks Steven __________________________________________________ Do You Yahoo!? Yahoo! Autos - Get free new car price quotes http://autos.yahoo.com

1 0

Preliminary patch for Bug 675 - LVM_SCROLL
by Guy L. Albertelli July 19, 2002

July 19, 2002

The attached patch implements LVM_SCROLL. While not many programs use it, implementing it required changing how scrollbars were handled for all of Listview. This new implementation is much more like the native. The scrollbars mimic the range of pixels that they cover. This allows the specification of a pixel level scroll which is critical to LVM_SCROLL. The only exception (and the native does this also) is for LVS_LIST displays where the scrollbar is related to the number of columns to be displayed. Before submitting to wine-patches, I am interested in feedback and any discrepancies with how the native controls work (-dll commctrl,comctl32=n). Thanks, Guy Albertelli <<galberte(a)neo.lrun.com>>

1 0

Re: wine/files dos_fs.c smb.c smb.h
by Rein Klazes July 19, 2002

July 19, 2002

On Thu, 18 Jul 2002 22:17:19 -0500, you wrote: > > Log message: > Mike McCormack <mikem(a)codeweavers.com> > First go at reading directories on public SMB shares. > > Patch: http://cvs.winehq.com/patch.py?id=1027048639509081130455346 This breaks Forte Agent, looping endlessly around a FindNextFileA(). It is caused by the SetLastError(ERROR_NO_MORE_FILES) that is lost in the final HeapUnlock call. A quick fix is attached. files/ : dos_fs.c Repair SetLastError in FindNextFileA. Rein. -- Rein Klazes rklazes(a)xs4all.nl

1 0

Re: Errors running winapi_extract --stub-statistics
by Gregg Mattinson July 19, 2002

July 19, 2002

I think this is because the module name is no longer in the .spec file. Lines 71-73 of winapi_extract look for "name" in the .spec file, which no longer exists. Should we be using the file name as the module name? Gregg Mattinson Co-op Developer Sun Microsystems of Canada >From: Francois Gouget <fgouget(a)free.fr> >To: wine-devel(a)winehq.com >cc: Patrik Stridvall <ps(a)leissner.se> >Subject: Errors running winapi_extract --stub-statistics >Date: Fri, 19 Jul 2002 00:19:15 -0700 (PDT) > > >Hi, > > >I was trying to update the API statistics but I got the following >errors: > >$ ./tools/winapi/winapi_extract --stub-statistics >Use of uninitialized value in hash element at >./tools/winapi/winapi_extract line 97. >Use of uninitialized value in hash element at >./tools/winapi/winapi_extract line 97. >... > >There is quite a few of these. So I tentatively modified the line as >follows: > $module2entries{$module} = $entries if (defined $module); > >But then I got: >$ ./tools/winapi/winapi_extract --stub-statistics >controls/button.c (file 1 of 796)Not a HASH reference at >tools/winapi_check/winapi_parser.pm line 32. > >So obviously there is another problem or my previous 'fix' was not so >good. Any ideas? > > >-- >Francois Gouget fgouget(a)free.fr http://fgouget.free.fr/ > In a world without fences who needs Gates? >

1 0

Patch for memory-mapping problem -- See VirtualQuery post
by Max July 19, 2002

July 19, 2002

Well, I took off the dirty hack on virtual.c and put inside something that seems to me much more like a real solution. Took off in executable mapping VPROT_WRITE and VPROT_WRITECOPY flags and added VPROT_EXEC; that looks much more close to win2k mapping. The only difference now is in Type field, wine maps as MEM_PRIVATE and win2k as MEM_IMAGE (I guess that's the right one, but no time now to check. Here the attached patch.... any comment appreciated :-) Regards Max

1 0

Errors running winapi_extract --stub-statistics
by Francois Gouget July 19, 2002

July 19, 2002

Hi, I was trying to update the API statistics but I got the following errors: $ ./tools/winapi/winapi_extract --stub-statistics Use of uninitialized value in hash element at ./tools/winapi/winapi_extract line 97. Use of uninitialized value in hash element at ./tools/winapi/winapi_extract line 97. ... There is quite a few of these. So I tentatively modified the line as follows: $module2entries{$module} = $entries if (defined $module); But then I got: $ ./tools/winapi/winapi_extract --stub-statistics controls/button.c (file 1 of 796)Not a HASH reference at tools/winapi_check/winapi_parser.pm line 32. So obviously there is another problem or my previous 'fix' was not so good. Any ideas? -- Francois Gouget fgouget(a)free.fr http://fgouget.free.fr/ In a world without fences who needs Gates?

1 0