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

Re: How to make a stub dll ?

by Sylvain Petreolle

I will do it monday (I don't have access to my computer now). > Well, at least submit your work about D3D 8 > stubbing. Let's hope that it > will once contain 'real' stuff :-) Sure it will :) > > And well, why would anyone need a bug to submit a > patch ? We are not using > yet such an advanced software process ! No, but we could create a metabug for d3d implementation. d3d8.dll contains a lot of calls to sub d3d8xxxxx dlls that we will emulate. > Yeah, but when you look at 'dlls/Makefile.in' to add > the new directory, you > will see that it's an auto-generated file => you > will use make_dlls :-) > 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. ___________________________________________________________ Do You Yahoo!? -- Une adresse @yahoo.fr gratuite et en français ! Yahoo! Mail : http://fr.mail.yahoo.com

June 21, 2002

GDI patch for Acad. Please feedback !

by P. Christeas

Below there is a patch which makes AutoCAD R14 work under wine. The patch involves (patch1.diff) some calculation for Bitmaps drawn in negative-y-axis (i.e. bottom-left is 0,0) windows. I just found that solution by trying. Therefore, the patch, may cause trouble to other apps (mostly drawing - CAD ones). *Please* try this patch and report if any app (just any!) mis-behaves with it. If you have any other solution for the same point, please tell me. The file 'patch2.diff' has one more suggestion, concerning NULL object handles. This should only introduce a minor speedup and suppress a warning message, which made gdi debugging little hard.

June 21, 2002

Re: How to make a stub dll ?

by Lionel Ulmer

> I now have a box saying that Direct3D 8.0 isn't > detected instead of a crash and my stub function > reports it correctly on the d3d debug channel. Nice :-) > I see there's no bug for d3d implementation. > Should I create it / submit diffs ? Well, at least submit your work about D3D 8 stubbing. Let's hope that it will once contain 'real' stuff :-) And well, why would anyone need a bug to submit a patch ? We are not using yet such an advanced software process ! > > 5) run make_dlls in the dlls directory > it isn't in the doc now (grep make_dlls * returns > nothing in wine/documentation) Yeah, but when you look at 'dlls/Makefile.in' to add the new directory, you will see that it's an auto-generated file => you will use make_dlls :-) > > 6) add a 'dlls/d3d8/Makefile' line in configure.ac > > and regenerate configure > needs autoconf 2.53 & automake 1.6, not available Not available ? I downloaded them from ftp.gnu.org and they work fine. Lionel -- Lionel Ulmer - http://www.bbrox.org/

June 20, 2002

Re: Notes R4 redraw problem

by Francois Gouget

On 20 Jun 2002, Vincent Béron wrote: [...] > What's the fundamental difference between WINE_ROOT/WINEDLLPATHS and > WINE_PREFIX? WINE_ROOT is just an environment variable I use so that I don't have to type the full path to the Wine sources over and over. It's even more useful in my wenv script. WINEDLLPATH tells Wine where to find the builtin Wine dlls. Here /home/fgouget/wine/dlls WINE_PREFIX tells Wine where to find its configuration file. By default this is $HOME/.wine -- Francois Gouget fgouget(a)free.fr http://fgouget.free.fr/ If it stinks, it's chemistry. If it moves, it's biology. If it does not work, It's computer science.

June 20, 2002

Re: Have regapi installed

by Andriy Palamarchuk

Bill, do you have any specific reasons to install regapi? It is obsoleted by regedit and (IMHO) should be removed from the tree. Andriy __________________________________________________ Do You Yahoo!? Yahoo! - Official partner of 2002 FIFA World Cup http://fifaworldcup.yahoo.com

June 20, 2002

Re: Notes R4 redraw problem

by Vincent Béron

Le jeu 20/06/2002 à 14:57, Francois Gouget a écrit : > On 20 Jun 2002, Vincent Béron wrote: > [...] > > Wouldn't it be possible to just install it in a scratch directory in the > > home dir or in /tmp? That's what I do now for test builds, and it works > > pretty well. I just use a wrapper script (yes, another) to wine, which > > sets up the environment variables needed for it to work (PATH, > > LD_LIBRARY_PATH, etc.). > > You don't even need to install it at all. I have never installed Wine on > my system. The installation part makes it easy afterwards to install system-wide when it works. But I agree you could do it after the testing, or not at all. > > What I do is something along the lines of: > > export WINE_ROOT=/home/fgouget/wine/wine > export LD_LIBRARY_PATH="$WINE_ROOT:$LD_LIBRARY_PATH" > export PATH="$WINE_ROOT:$WINE_ROOT/server:$WINE_ROOT/tools:$PATH" > export WINEDLLPATH="$WINE_ROOT/dlls" > > Actually I have a script that does the above and which I use to pick one > Wine tree or another (". wenv wine" or ". wenv wine2", etc.). What's the fundamental difference between WINE_ROOT/WINEDLLPATHS and WINE_PREFIX? > > I also have a ~/.wine directory with a fake_windows configuration that I > reuse from one time to the next. I also have it tarred so I can in > essence "rm -rf ~/.wine && tar xvfz dotwine.tar.gz". I also have an orig.tar.gz which is essentially the same thing, although it also archives the etc dir. When I have a new test tree, I untar it, then modify about 4 paths in it (I could use a sed script during the installation, but I'm too lazy for now to sit down and write it). Vincent

June 20, 2002

Re: Notes R4 redraw problem

by Francois Gouget

On 20 Jun 2002, Vincent Béron wrote: [...] > Wouldn't it be possible to just install it in a scratch directory in the > home dir or in /tmp? That's what I do now for test builds, and it works > pretty well. I just use a wrapper script (yes, another) to wine, which > sets up the environment variables needed for it to work (PATH, > LD_LIBRARY_PATH, etc.). You don't even need to install it at all. I have never installed Wine on my system. What I do is something along the lines of: export WINE_ROOT=/home/fgouget/wine/wine export LD_LIBRARY_PATH="$WINE_ROOT:$LD_LIBRARY_PATH" export PATH="$WINE_ROOT:$WINE_ROOT/server:$WINE_ROOT/tools:$PATH" export WINEDLLPATH="$WINE_ROOT/dlls" Actually I have a script that does the above and which I use to pick one Wine tree or another (". wenv wine" or ". wenv wine2", etc.). I also have a ~/.wine directory with a fake_windows configuration that I reuse from one time to the next. I also have it tarred so I can in essence "rm -rf ~/.wine && tar xvfz dotwine.tar.gz". -- Francois Gouget fgouget(a)free.fr http://fgouget.free.fr/ In theory, theory and practice are the same, but in practice they're different.

June 20, 2002

Re: Notes R4 redraw problem

by Vincent Béron

Le jeu 20/06/2002 à 14:16, Ian Pilcher a écrit : > Tony Lambregts wrote: > > > > Binary searching is very efficient. As far as the theory goes I am > > confident that a binary search for the bug should take 12 or or less > > itinerations. The problem with this is that sometimes there are other > > regressions that mask the original regression. I do not have lotus notes > > (any version) but I do know about regression testing using CVS. > > > > Tony - > > If a binary search needs to be done, I'll do it. (I can build and > install Wine in about 10 minutes on my system, so the compile time won't > be a problem.) I am hoping to avoid it, however, because I'll have to > carve out and set up a complete test partition; I always use RPM to > install software on my production systems, and I don't want to spend the > time tweaking the SPEC file for each build. > > Hopefully someone will be able to make sense of a message trace. > -- > ======================================================================== > Ian Pilcher pilchman(a)attbi.com > ======================================================================== > > Wouldn't it be possible to just install it in a scratch directory in the home dir or in /tmp? That's what I do now for test builds, and it works pretty well. I just use a wrapper script (yes, another) to wine, which sets up the environment variables needed for it to work (PATH, LD_LIBRARY_PATH, etc.). In fact, I attach it here if you'd need it (or if somebody else needs it). Of course, you also need to tell ./configure --prefix=/where/you/want/it, then it's a normal make depend, make, make install. The first time, you'll also need to modify your wine.conf and install the basic registry. And when you're finished with it, rm -rf (or F8 in mc). Vincent

June 20, 2002

Re: Notes R4 redraw problem

by Ian Pilcher

Tony Lambregts wrote: > > Binary searching is very efficient. As far as the theory goes I am > confident that a binary search for the bug should take 12 or or less > itinerations. The problem with this is that sometimes there are other > regressions that mask the original regression. I do not have lotus notes > (any version) but I do know about regression testing using CVS. > Tony - If a binary search needs to be done, I'll do it. (I can build and install Wine in about 10 minutes on my system, so the compile time won't be a problem.) I am hoping to avoid it, however, because I'll have to carve out and set up a complete test partition; I always use RPM to install software on my production systems, and I don't want to spend the time tweaking the SPEC file for each build. Hopefully someone will be able to make sense of a message trace. -- ======================================================================== Ian Pilcher pilchman(a)attbi.com ========================================================================

June 20, 2002

Re: [Bug 628] Changed - HELP-WIndows app. fail to run

by Andriy Palamarchuk

--- wqe dsf liolio <rhlinux(a)rediffmail.com> wrote: > Hello, > I haven't tried the suggestion yet, but will be > doing it > shortly. > Please let me know how to add comments to the > already reported > bug. > KK. Each bug notification message has a hyperlink to the bug details page. In this case it is : > On Thu, 06 Jun 2002 wine-bugs(a)winehq.com wrote : > >http://bugs.winehq.com/show_bug.cgi?id=628 Go there, you should be able to add comments and submit them with "Submit" button. If you don't see the button under the "Additional Comments" text field this means you have to login. Let me know if you still can't add comments. Andriy __________________________________________________ Do You Yahoo!? Yahoo! - Official partner of 2002 FIFA World Cup http://fifaworldcup.yahoo.com

June 20, 2002

How do I obtain GUIDs for socket transports?

by Martin Wilck

Hi, I am currently trying to create a "clean" implementation of WSAEnumProtocols and friends. Eventually WSASocket, WSADuplicateSocket, getsockopt etc. will support the WSAPROTOCOL_INFO structs correctly. Here is a question: Windows has a GUID assigned to each protocol. "uuids.h" contains a large number of wine-UUIDs - how were these constructed? How should I build mine? We are not talking about random GUIDs here. My NT box shows e.g. e70f1aa0-ab8b-11cf-8ca300805f48a192 for MS TCP/IP, 11058240-be47-11cf-95c800805f48a192 for MS IPX, 2c3b17a0-c6df-11cf-95c800805f48a192 for MS Appletalk. Should I simply copy these for wine, or change them somehow? Martin -- Martin Wilck Phone: +49 5251 8 15113 Fujitsu Siemens Computers Fax: +49 5251 8 20409 Heinz-Nixdorf-Ring 1 mailto:Martin.Wilck(a)Fujitsu-Siemens.com D-33106 Paderborn http://www.fujitsu-siemens.com/primergy

June 20, 2002

Re: [Bug 628] Changed - HELP-WIndows app. fail to run

by Tony Lambregts

wqe dsf liolio wrote: > Hello, > I haven't tried the suggestion yet, but will be doing it shortly. > Please let me know how to add comments to the already reported bug. > KK. > > > [snip] > >> http://bugs.winehq.com/show_bug.cgi?id=628 > [snip] click on the link The first line of a notifcation is the link to the bug report where you can enter your comments. Tony Lambregts

June 20, 2002

Re: Notes R4 redraw problem

by Ian Pilcher

Daniel Schwarz wrote: > > Ian, do you know if this bug occurred with older builds of Wine, or can > you state that it definitely did not occur? > Dan - It's definitely a regression. I was a fairly regular user of Notes R4 under Wine until about 12-18 months ago. (I finally got broadband access and began working from home at that time. Until very recently, however, my employers VPN client was Windows-only.) This problem was definitely not present back then. I'm going to try the message tracing suggestion first. Is there a better way than '--debugmsg +message'? -- ======================================================================== Ian Pilcher pilchman(a)attbi.com ========================================================================

June 20, 2002

Re: wine can't find regedit.exe.so when it is not on the path

by Rob Browning

Alexandre Julliard <julliard(a)winehq.com> writes: >> wineinstall fails because wine can't find >> regedit.exe.so. >> Obviously, the wine source directory does not exist in >> path of just created config file. > > I'd say that wineinstall should run the installed wine binaries, not > the ones from the source dir. I recently tried the newish build infrastructure (using wineinstall), and overall I liked it, but I wasn't able to use it without some modifications. That's because I install wine into /usr/local/opt/wine, I never install it as root, and I don't want it tinkering with my ld.so.conf. I want to handle getting the right bits in my PATH and LD_LIBRARY_PATH myself. However, I *would* like to let wineinstall handle all the rest. So I was wondering if there would be any interest in a patch to wineinstall that would allow this kind of install as an option. Perhaps something like this: tools/wineinstall --not-root --prefix=/usr/local/opt/wine Even easier would be to just allow envt var overrides rather than coding for command line options. i.e. WINEINSTALL_SU=no WINEINSTALL_PREFIX=/usr/local/opt/wine tools/wineinstall I could also output a final message would be output when anyone specifies a prefix that says something like You have used a prefix other than the default. You will need to make sure your PATH includes ... and your LD_LIBRARY_PATH includes ... If you don't want to have to do this, please don't specify WINEINSTALL_PREFIX. Thanks -- Rob Browning rlb @defaultvalue.org, @linuxdevel.com, and @debian.org Previously @cs.utexas.edu GPG=1C58 8B2C FB5E 3F64 EA5C 64AE 78FE E5FE F0CB A0AD

June 20, 2002

Re: Notes R4 redraw problem

by Tony Lambregts

Ian Pilcher wrote: >dschwarz(a)bellatlantic.net wrote: > > >>Question: Once this happens, can you drag the vertical pane rightwards and >>redisplay the folder view? Or is it gone until you quit/restart Notes? >> >> >> > >It's easy to resize the pane so that the folder view is visible again, >so this isn't a major problem. I should probably also have mentioned >that this didn't happen with older versions of Wine (circa a year ago). > > > I you want to do a set of regression tests I estimate that it should take no more then 12 itinerations to find the day of patch that broke it . Thats the theory anyway, but it might not work <g> Tony Lambregts

June 19, 2002