July 2003 - Wine-devel - list.winehq.org

RE: [RESENT] Always print version on startup
by PETREOLLE Sylvain 01 Jul '03

01 Jul '03

I think this can be avoided. Let use stderr to print this, this way you can use redirections with every program you want. I will add one more remark (!) : I dont know many programs that expect "Wine version" in their output. If they are disturbed by this, this is a huge design problem. -----Message d'origine----- De : Francois Gouget [mailto:fgouget(a)free.fr] Envoyé : mardi 1 juillet 2003 23:57 À : wine-devel(a)winehq.org Objet : Re: [RESENT] Always print version on startup Let me object too then. This patch will break all scripts that use regedit to export part of the registry to stdout. It will also break all applications that run 'uninstaller --list' and then parse the output to determine what Windows applications are installed. It will also break all Winelib applications that print anything to stdout that then needs to be parsed by a script or some other program. Of course we have each of these cases in CrossOver. Then it will break all Windows application that analyze the output of another Windows application. For instance I expect Visual C++'s nmake+cl will not work anymore (as well as other make+compiler cases). Even compiling from the Visual C++ IDE may stop working. Finally it's a matter of what Wine is about. Wine is about transparently running Windows applications. Wine itself should be as invisible as possible, it should just get out of the way. We went out of our way (well iirc Alexandre did most of that work) to avoid having the Wine command line options interfer with the application's command line options. Printing garbage (from the Windows application point of vue) on stdout would be a step backwards. -- Francois Gouget fgouget(a)free.fr http://fgouget.free.fr/ War doesn't determine who's right. War determines who's left.

2 1

Re: [RESENT] Always print version on startup
by Francois Gouget 01 Jul '03

01 Jul '03

On Tue, 1 Jul 2003, Uwe Bonnes wrote: > Changelog: > misc/options.c: OPTIONS_ParseOptions > Always print out version > > > The recent posting about the Altere quartus installer not working seems to > ^^^^^^ > Altera > > be caused by an old version used. The poster shows the startup and > > messages. If the version was printed by default, it could help debugging > > Only Marcus objected, but didn't react on my remarks. [...] Let me object too then. This patch will break all scripts that use regedit to export part of the registry to stdout. It will also break all applications that run 'uninstaller --list' and then parse the output to determine what Windows applications are installed. It will also break all Winelib applications that print anything to stdout that then needs to be parsed by a script or some other program. Of course we have each of these cases in CrossOver. Then it will break all Windows application that analyze the output of another Windows application. For instance I expect Visual C++'s nmake+cl will not work anymore (as well as other make+compiler cases). Even compiling from the Visual C++ IDE may stop working. Finally it's a matter of what Wine is about. Wine is about transparently running Windows applications. Wine itself should be as invisible as possible, it should just get out of the way. We went out of our way (well iirc Alexandre did most of that work) to avoid having the Wine command line options interfer with the application's command line options. Printing garbage (from the Windows application point of vue) on stdout would be a step backwards. -- Francois Gouget fgouget(a)free.fr http://fgouget.free.fr/ War doesn't determine who's right. War determines who's left.

1 0

Re: [DSHOW-03a] Fix Compile Problems On Newer Compilers
by Maxime Bellengé 01 Jul '03

01 Jul '03

On my redhat 9, this patch doesn't work, I have to move the two defines at the very begining of the filtermapper.c file to make it work. Max On Tue, 2003-07-01 at 19:23, Robert Shearman wrote: > Hi, > > This should fix the problems a number of people were having when compiling > after the commiting of my DSHOW-02 patch. > > Changelog: > Fix compilation due to nameless structs and unions > > Rob -- Maxime Bellengé <maxime.bellenge(a)laposte.net>

1 0

can't compile current cvs - filtermapper
by hatky 01 Jul '03

01 Jul '03

current cvs, completly clean: make[2]: Entering directory `/home/hatky/wine-source/wine/dlls/quartz' gcc -c -I. -I. -I../../include -I../../include -D_REENTRANT -fPIC -D__WINESRC__ -Wall -mpreferred-stack-boundary=2 -gstabs+ -Wpointer-arith -g -O2 -o filterm apper.o filtermapper.c filtermapper.c: In function `FM2_WriteFilterData': filtermapper.c:365: structure has no member named `u' filtermapper.c:370: structure has no member named `u' filtermapper.c:373: structure has no member named `u' filtermapper.c:375: structure has no member named `u' filtermapper.c:376: structure has no member named `u' filtermapper.c:379: structure has no member named `u' filtermapper.c:382: structure has no member named `u' filtermapper.c: In function `FM2_ReadFilterData': filtermapper.c:513: structure has no member named `u' filtermapper.c:515: structure has no member named `u' filtermapper.c: In function `FM2_DeleteRegFilter': filtermapper.c:602: structure has no member named `u' filtermapper.c:605: structure has no member named `u' filtermapper.c:606: structure has no member named `u' filtermapper.c:608: structure has no member named `u' filtermapper.c:610: structure has no member named `u' filtermapper.c:611: structure has no member named `u' filtermapper.c:613: structure has no member named `u' filtermapper.c: In function `FilterMapper2_EnumMatchingFilters': filtermapper.c:910: structure has no member named `u' filtermapper.c:912: structure has no member named `u' make[2]: *** [filtermapper.o] Error 1 make[2]: Leaving directory `/home/hatky/wine-source/wine/dlls/quartz' make[1]: *** [quartz] Error 2 make[1]: Leaving directory `/home/hatky/wine-source/wine/dlls' make: *** [dlls] Error 2 Compilation failed, aborting install. any ideas? did not find anything at the bord already, Hatky. __________________________________ Do you Yahoo!? SBC Yahoo! DSL - Now only $29.95 per month! http://sbc.yahoo.com

1 0

Compile Error in CVS as of 7/1/03 at 7:14am MST
by Kevin DeKorte 01 Jul '03

01 Jul '03

gcc -c -I. -I. -I../../include -I../../include -D_REENTRANT -fPIC -D__WINESRC__ -Wall -mpreferred-stack-boundary=2 -gstabs+ -Wpointer-arith -g -O2 -o filtermapper.o filtermapper.c filtermapper.c: In function `FM2_WriteFilterData': filtermapper.c:365: structure has no member named `u' filtermapper.c:370: structure has no member named `u' filtermapper.c:373: structure has no member named `u' filtermapper.c:375: structure has no member named `u' filtermapper.c:376: structure has no member named `u' filtermapper.c:379: structure has no member named `u' filtermapper.c:382: structure has no member named `u' filtermapper.c: In function `FM2_ReadFilterData': filtermapper.c:513: structure has no member named `u' filtermapper.c:515: structure has no member named `u' filtermapper.c: In function `FM2_DeleteRegFilter': filtermapper.c:602: structure has no member named `u' filtermapper.c:605: structure has no member named `u' filtermapper.c:606: structure has no member named `u' filtermapper.c:608: structure has no member named `u' filtermapper.c:610: structure has no member named `u' filtermapper.c:611: structure has no member named `u' filtermapper.c:613: structure has no member named `u' filtermapper.c: In function `FilterMapper2_EnumMatchingFilters': filtermapper.c:910: structure has no member named `u' filtermapper.c:912: structure has no member named `u' make[2]: *** [filtermapper.o] Error 1 make[2]: Leaving directory `/home/kdekorte/cvs/wine/dlls/quartz' make[1]: *** [quartz] Error 2 make[1]: Leaving directory `/home/kdekorte/cvs/wine/dlls' make: *** [dlls] Error 2

1 0

Nameless struct in dlls/quartz/filtermapper.c
by Uwe Bonnes 01 Jul '03

01 Jul '03

Hallo, with an CVS update this morning, compiling in dlls/quartz/ gives: gcc -c -I. -I. -I../../include -I../../include -D_REENTRANT -fPIC -D__WINESRC__ -Wall -mpre ferred-stack-boundary=2 -gstabs+ -Wpointer-arith -g -O2 -o filtermapper.o filtermapper.c filtermapper.c: In function `FM2_WriteFilterData': filtermapper.c:365: structure has no member named `u' filtermapper.c:370: structure has no member named `u' ... These problems with nameless structs pop up on and off... 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: wine/dlls/quartz quartz_private.h main.c filte ...
by Robert Shearman 01 Jul '03

01 Jul '03

> I updated from cvs an get an error I apologise for this. I will fix this when I next get access to my computer (it will be a fair few hours), unless someone beats me to it. It is a case of putting the following defines before strmif.h is included: #define NONAMELESSUNION #define NONAMELESSSTRUCT Rob

1 0

Re: wine/dlls/quartz quartz_private.h main.c filte ...
by Marcelo Duarte 01 Jul '03

01 Jul '03

I updated from cvs an get an error: gcc -c -I. -I. -I../../include -I../../include -D_REENTRANT -fPIC -D__WINESRC__ -Wall -mpreferred-stack-boundary=2 -gstabs+ -Wpointer-arith -g -O2 -o filtermapper.o filtermapper.c filtermapper.c: In function `FM2_WriteFilterData': filtermapper.c:365: structure has no member named `u' filtermapper.c:370: structure has no member named `u' filtermapper.c:373: structure has no member named `u' filtermapper.c:375: structure has no member named `u' filtermapper.c:376: structure has no member named `u' filtermapper.c:379: structure has no member named `u' filtermapper.c:382: structure has no member named `u' filtermapper.c: In function `FM2_ReadFilterData': filtermapper.c:513: structure has no member named `u' filtermapper.c:515: structure has no member named `u' filtermapper.c: In function `FM2_DeleteRegFilter': filtermapper.c:602: structure has no member named `u' filtermapper.c:605: structure has no member named `u' filtermapper.c:606: structure has no member named `u' filtermapper.c:608: structure has no member named `u' filtermapper.c:610: structure has no member named `u' filtermapper.c:611: structure has no member named `u' filtermapper.c:613: structure has no member named `u' filtermapper.c: In function `FilterMapper2_EnumMatchingFilters': filtermapper.c:910: structure has no member named `u' filtermapper.c:912: structure has no member named `u' make[2]: ** [filtermapper.o] Erro 1 make[2]: Leaving directory `/home/mduarte/wine/dlls/quartz' make[1]: ** [quartz] Erro 2 make[1]: Leaving directory `/home/mduarte/wine/dlls' make: ** [dlls] Erro 2 Em Ter 01 Jul 2003 01:29, Alexandre Julliard escreveu: > ChangeSet ID: 8645 > CVSROOT: /home/winehq/opt/cvs-commit > Module name: wine > Changes by: julliard(a)wine.codeweavers.com 2003/06/30 23:29:48 > > Modified files: > dlls/quartz : quartz_private.h main.c filtergraph.c > Makefile.in > Added files: > dlls/quartz : filtermapper.c enummoniker.c > > Log message: > Robert Shearman <R.J.Shearman(a)warwick.ac.uk> > Implemented IFilterMapper2. > > Patch: > http://cvs.winehq.com/patch.py?root=/home/winehq/opt/cvs-commit&id=8645 > > Old revision New revision Changes Path > 1.7 1.8 +3 -11 wine/dlls/quartz/quartz_private.h > 1.24 1.25 +2 -0 wine/dlls/quartz/main.c > 1.2 1.3 +12 -0 wine/dlls/quartz/filtergraph.c > 1.26 1.27 +3 -1 wine/dlls/quartz/Makefile.in > Added 1.1 +0 -0 wine/dlls/quartz/filtermapper.c > Added 1.1 +0 -0 wine/dlls/quartz/enummoniker.c -- Marcelo Duarte

1 0

Deficient implementation of SetFileTime or with stub GetSetFileTimestamp
by Marcelo Duarte 01 Jul '03

01 Jul '03

Hi, Installing a system of the company in the Wine, I found a probable error, that is the following one: when the archives go being unpacked, them they are created in the directory with the current date and would have to be created with the same date of the compact file (as it is made in windows). Thus, in the next step, an update is made verifying the dates it of the files. However, in the Wine, it gives wrong for cause of this difference in the dates. I believe that I can myself make a patch, however as I still do not have experience in debugging Wine, he would like a suggestion of as to find the problem, because I do not know if the bug is in lzexpand (LZCopy) or in SetupApi (perhaps because of stub GetSetFileTimestamp) or because in SetFileTime "const FILETIME * lpCreationTime" simply is ignored, and in server, the function "static int set_file_time(obj_handle_t handle, time_t access_time, time_t write_time) " must not have more one parameter? Manually I make the system to install and run, however I want it running like in Windows. Thanks in advance. -- Marcelo Duarte

1 0

Re: Various problems- it just gets worse!
by Shachar Shemesh 01 Jul '03

01 Jul '03

Duane Clark wrote: > Shachar Shemesh wrote: > >> Duane Clark wrote: >> >> >>> I do understand the aversion to scripts, but the wineinstall one is >>> fairly mild. I've been using Wine for more than 3 years and have >>> even done some development on it (I wish I had more time for that... >>> sigh). Still, if I want to test with a clean install, I delete or >>> move the old windows and .wine directories and just use wineinstall >>> to create fresh versions. Because, well, it just works so well and >>> is a lot easier than doing the configuration by hand. >> >> >> >> My beef with wineinstall is that there is no way to isolate the >> configuration creation part from the install part. I think we should >> have a script that does only the configuration, fake root and >> registry creation. This way, the script can be bundled with the >> package and placed on the destination machine. such a script should >> be runnable without the sources dir (maybe by specifying the location >> of the template registry and such via command line, maybe created by >> ./configure). >> >> This way, the order CAN be "./configure, make depend, make, make >> install, createconf" for normal install. Packagers can then do >> "./configure --templatedir=/usr/share/wine --prefix=/usr (etc) && >> make depend && make && TMPDIR=/tmp/winerpmroot make install", and run >> "createconf" from the "postinstall" section of the RPM sort of thing. >> >> Opinions? > > > It sounds like you are looking for something useful for binary > distributions. I think that wineinstall would normally only be used by > someone who is downloading and installing their first source > distribution. Not entirely true either. I often find myself seeking to restore the config. Also, what about someone who downloaded the sources and installed wine system wide, and then wants to enable wine for a second user? > So I think wineinstall is appropriate for that. > > For binary distributions, I think a GUI tool is more appropriate. And > there exists a tool, winesetuptk, for that purpose. It is a long time > since I tried that tool, but as I recall, it was fairly easy to use. > I disagree, even had I been looking for something useful for a binary distribution. A GUI tool would be the worst solution possible. winesetuptk is excellent for an end user trying to tweak his own configuration. Something that runs while rpm -i (or dpkg -i) cannot depend on GUI, for it would prevent it from running while installing the system. Shachar -- Shachar Shemesh Open Source integration consultant Home page & resume - http://www.shemesh.biz/

1 0