June 2002 - Wine-Devel - list.winehq.org

Re: Patch 1/2 for SmartForms Installer: PROFILE_GetString incorrect return value

by Andreas Mohr

On Sat, Jun 22, 2002 at 12:18:07AM -0400, Joshua Thielen wrote: > This patch fixes an error encountered in the Smart Forms installer > (somewhat related to bug 575). > > When calling GetPrivateProfileString with the second parameter as NULL, > all of the keys under the specified section are copied to the buffer. > The return value should be the number of characters copied minus the > last null terminator. Since the keys are copied as null-terminated > strings with the last string terminated with two null characters, strlen > cannot be used to determine the number of characters copied. This patch > corrects PROFILE_GetString for this case. Hmm, hmm. Could this happen to fix bug #724 also ? Word97 etc. broke when I handled the "default" case in PROFILE_GetString(). As the log files received didn't show relay lines, only profile lines, I haven't been able to fix it. But this sounds like a reasonable cause for it... -- Andreas Mohr Stauferstr. 6, D-71272 Renningen, Germany Tel. +49 7159 800604 http://mohr.de.tt

June 22, 2002

Re: Patch 2/2 for SmartForms Installer: GetModuleFileNameW called by 16-bit task.

by Dmitry Timoshkov

"Joshua Thielen" <thielen(a)netprince.net> wrote: > - if ((wm = MODULE32_LookupHMODULE( hModule ))) > + if(!hModule && !(NtCurrentTeb()->tibflags &= TEBF_WIN32)) { ^^ Apparently '&=' above must be replaced by simple '&' -- Dmitry.

June 22, 2002

Question about printer support in wine.

by Joshua Thielen

Hello everybody, I'm trying to set up my printer in wine in order to test SmartForms and LogicWorks (they both are helplessly out of order without a default printer). The documentation says that if you have CUPS installed, all printing options are autodetected regardless of .ini or registry entries. I have cups, libcups, and libcups-devel 1.1.0 installed and I'm running Mandrake 8.1. Configure is able to find the CUPS files and enables CUPS printing support in wine. But when I run either SmartForms or LogicWorks, it says that a default printer could not be found. Is the documentation correct, or am I missing something? I'm sorry, I probably should have posted this question to a different list, but I'm too lazy to go elsewhere. TIA, Josh

June 22, 2002

Re: GLU32 patch (yep, again)

by Alexandre Julliard

Vincent Béron <vberon(a)mecano.gme.usherb.ca> writes: > glu32 can now be built with a libGLU 1.1 or 1.2. If a 1.1 version is > used, some functions only issue a WARN. It should become more and more > rare to only have a libGLU 1.1. > > If there's still something preventing this from being applied, please > let me know. It's not clear to me what we gain by using the header file. If you need to check for the presence of some specific functions you should do a configure check for these, instead of using a GLU_VERSION define. Even better would be to load them dynamically using dlsym, like we do for libfreetype and libXrender. -- Alexandre Julliard julliard(a)winehq.com

June 21, 2002

Re: Libwine brokeness on Mingw

by Alexandre Julliard

"Steven Edwards" <Steven_Ed4153(a)yahoo.com> writes: > How should I fix this in the makefiles? I don't need config.c at all and > I will need to do something similer for the win16 code in other dlls I'd suggest fixing config.c instead of not using it. There will likely be other things added to that file that you might need too. -- Alexandre Julliard julliard(a)winehq.com

June 21, 2002

Re: Regression errors ...

by Andreas Mohr

On Fri, Jun 21, 2002 at 04:20:21PM +0100, Paul Millar wrote: > Hi, > > Last nights patch for DRIVE_FindDriveRoot: > http://cvs.winehq.com/patch.py?id=1024614786458303934502513 > caused two regression errors in dlls/kernel/tests/path.c at lines 212 and > 672 ("GetShortPathNAmeA did not return an 8.3 path" and "GetShortPathNameA > did not keep relative directory" respectively) Heh, there's nothing like a good regression test suite :-) OTOH I really don't want to know how much trouble this patch causes everywhere else, given that it even already triggers with the scarce few regression functions that we currently have. Somebody should probably have a second hard look at the possible effects of this severe change in file system functionality... > BTW, the TODO test (line 2435) in dlls/oleaut32/tests/vartest.c succeeds. > Should it be move from TODO to a real test? Yes, definitely ! That's most likely due to the tons of oleaut32 fixes yesterday. (sounds like that has been very good work, then...) BTW: I'm finally about to submit my first regression test, too. -- Andreas Mohr Stauferstr. 6, D-71272 Renningen, Germany Tel. +49 7159 800604 http://mohr.de.tt

June 21, 2002

Re: PATCH: documentation/samples/config

by Andreas Mohr

On Fri, Jun 21, 2002 at 12:14:37PM +0200, Johan Gill wrote: > Changelog: 48 was plain wrong as value for HELmargin. Removed the sample > EmulDriver appdefault since N is the default now, and an appdefault for > Y is not useful. Huh ?? I'm sure there are plenty of entries in there that are default, but a sample config file is supposed to be a *sample* file... How do you expect people to know about EmulDriver if you get plain rid of it ? EmulDriver is very important, BTW. Several games need a different setting for it. -- "My attitude is, everybody should try competing with Microsoft once in their life. Once." - Marc Andreessen, former Netscape lead employee, in a "browser wars" interview

June 21, 2002

Regression errors ...

by Paul Millar

Hi, Last nights patch for DRIVE_FindDriveRoot: http://cvs.winehq.com/patch.py?id=1024614786458303934502513 caused two regression errors in dlls/kernel/tests/path.c at lines 212 and 672 ("GetShortPathNAmeA did not return an 8.3 path" and "GetShortPathNameA did not keep relative directory" respectively) BTW, the TODO test (line 2435) in dlls/oleaut32/tests/vartest.c succeeds. Should it be move from TODO to a real test? Cheers, ---- Paul Millar

June 21, 2002

Libwine brokeness on Mingw

by Steven Edwards

This patch http://cvs.winehq.com/cvsweb/wine/library/Makefile.in.diff?r1=1.14&r2=1. 15 How should I fix this in the makefiles? I don't need config.c at all and I will need to do something similer for the win16 code in other dlls Thanks Steven "Every revolution was once a thought in one man's mind" - Ralph Waldo Emerson

June 21, 2002

Re: How to make a stub dll ?

by Vincent Béron

Le ven 21/06/2002 à 08:51, Sylvain Petreolle a écrit : > > > Not available ? I downloaded them from ftp.gnu.org > > and they work fine. > Sorry, forgot something in my sentence : > no rpms are available for redhat 7.3 now, I downloaded > the RawHide ones. There's an autoconf253-2.53-3 package from RH for RH7.3 (rather than autoconf-2.53), although I only see an automake15-1.5-2 package and no automake16 or automake-1.6. Also, I didn't know automake 1.6 was needed; indeed, I rebuilt my wine/configure last week with a standard RH7.3 box wrt those two packages without problems. > > ___________________________________________________________ > Do You Yahoo!? -- Une adresse @yahoo.fr gratuite et en français ! > Yahoo! Mail : http://fr.mail.yahoo.com > Vincent

June 21, 2002