April 2006 - Wine-Devel - list.winehq.org

Re: winebrowser: Use user's preferred browser
by Jeremy White April 17, 2006

April 17, 2006

> + "kfmclient exec,gnome-open,kmail,mozilla-thunderbird,thunderbird,evolution"; A Gnome user will be offended if you spool up the entire KDE environment just to launch a URL; a KDE user will be similarly offended by gnome-open. Further, if I'm a Gnome user, I may have KDE on my system, and kfmclient exec may work, but it's likely to not be the browser I *really* want used. The standard that's developing is that you don't invoke a KDE program unless KDE_FULL_DESKTOP is set and you don't invoke a Gnome session unless GNOME_DESKTOP_SESSION_ID is set. Cheers, Jeremy

5 5

Re: winspool/tests: fix broken test for GetPrinterDriver
by Dmitry Timoshkov April 17, 2006

April 17, 2006

"Detlef Riekenberg" <wine.dev(a)web.de> wrote: > @@ -939,36 +939,36 @@ static void test_GetPrinterDriver(void) > > for (level = -1; level <= 7; level++) > { > - SetLastError(0xdeadbeef); > + SetLastError(MAGIC_DEAD); Please don't change this, stick to the style used everywhere else, i.e. use 0xdeadbeef directly. > needed = (DWORD)-1; > ret = GetPrinterDriver(hprn, NULL, level, NULL, 0, &needed); > - ok(!ret, "level %d: GetPrinterDriver should fail\n", level); > - if (level >= 1 && level <= 6) > - { > - ok(GetLastError() == ERROR_INSUFFICIENT_BUFFER, "wrong error %ld\n", GetLastError()); > - ok(needed > 0,"not expected needed buffer size %ld\n", needed); > - } > - else > - { > - ok(GetLastError() == ERROR_INVALID_LEVEL, "wrong error %ld\n", GetLastError()); > - ok(needed == (DWORD)-1,"not expected needed buffer size %ld\n", needed); > - continue; > - } > + ok(!ret && ((GetLastError() == ERROR_INVALID_LEVEL) || > + (GetLastError() == ERROR_INSUFFICIENT_BUFFER) || > + (GetLastError() == ERROR_OUTOFMEMORY)), "%d: returned %d with %ld " \ > + "(expected '0' with: ERROR_INVALID_LEVEL, ERROR_INSUFFICIENT_BUFFER" \ > + "or ERROR_OUTOFMEMORY)\n", level, ret, GetLastError()); > + > + if (GetLastError() != ERROR_INSUFFICIENT_BUFFER) continue; > + > + ok(needed > 0,"%d: returned size %ld\n", level, needed); > + if (needed == 0) continue; I object to this change. The test needs a clear indication whether it's an invalid or a not supported parameter. Do not remove if {} else {}, that's an essential part of the test. A comment explaining the change would be also nice. > ret = GetPrinterDriver(hprn, NULL, level, buf, needed, &filled); > - ok(ret, "level %d: GetPrinterDriver error %ld\n", level, GetLastError()); > - ok(needed == filled, "needed %ld != filled %ld\n", needed, filled); > + ok( ret, "%d: returned %d with %ld (expected '!= 0')\n", > + level, ret, GetLastError()); "expected '!= 0'" is wrong, ret is BOOL, so "expected TRUE" or better leave it as it is now, error message already implies that the API has failed. > - ok(di_2->cVersion >= 0 && di_2->cVersion <= 3, "di_2->cVersion = %ld\n", di_2->cVersion); > + ok((di_2->cVersion >= 0 && di_2->cVersion <= 3) || > + (di_2->cVersion == 0x0400), "di_2->cVersion = %ld\n", di_2->cVersion); A comment explaining which windows version returns 0x400 (which contradicts MSDN) would be helpful. > ok(di_2->pName != NULL, "not expected NULL ptr\n"); > ok(di_2->pEnvironment != NULL, "not expected NULL ptr\n"); > ok(di_2->pDriverPath != NULL, "not expected NULL ptr\n"); > @@ -977,15 +977,15 @@ static void test_GetPrinterDriver(void) > > trace("cVersion %ld\n", di_2->cVersion); > trace("pName %s\n", di_2->pName); > - calculated += strlen(di_2->pName) + 1; > + calculated += lstrlenA(di_2->pName) + 1; If you want to replace strlen by lstrlenA (why?) that should be a separate patch. Same for below changes. -- Dmitry.

1 0

Solaris: Rundll322 exercising
by Robert Lunnon April 16, 2006

April 16, 2006

I would like to exercise rundll32 or similar (IE accessing a unix lib via the link in in .wine). This didn't work before because LD.SO refused to load a lib via a link with a colon in it's name. I wrote some code to fix it, but now process.c has been rearranged and the patch doesn't apply. It may well be that the new code works now but I need to test.

1 0

Re: abort if WINEDEBUG requests functionality that was disabled at configure time
by Mike Hearn April 16, 2006

April 16, 2006

On Sat, 15 Apr 2006 23:37:17 -0400, Mike Frysinger wrote: > if wine is built with --disable-trace or --disable-debug, then using WINEDEBUG > will accept the respective options without complaining and without actually > showing any useful information Why would anybody build without tracing anyway? I would be *very* surprised if it led to any noticable/real world performance improvement. Rather than do this why not just restore your packages configure switches to the upstream defaults?

5 7

Wine missing from repositories?
by Dee Ayy April 16, 2006

April 16, 2006

1) I installed wine at work from kubunu breezy repositories (standard and maybe universe and multiverse repositories are added) on wednesday 4/12/2006, yet on thursday 4/13/2006 I installed ubuntu breezy at home and definitely had standard, universe, multiverse, and even the *deb http://wine.sourceforge.net/apt/ binary/* repository added, yet could not see wine available for installation. Is Wine missing from repositories? 2) I thought I might be able to build wine from source for my AMD64 using the info on this page http://www.winehq.com/site/download-deb , but I seem to be getting a ubuntu compiler problem. But I am posting the error in case it is a winehq problem and this helps you (and attaching config.log): root(a)ubuntu:/home/user1/apt_sources# apt-get --build source wine Reading package lists... Done Building dependency tree... Done Need to get 13.6MB of source archives. Get:1 http://wine.sourceforge.net source/ wine 0.9.11~winehq1-1 (dsc) [1209B] Get:2 http://wine.sourceforge.net source/ wine 0.9.11~winehq1-1 (tar) [ 13.5MB] Get:3 http://wine.sourceforge.net source/ wine 0.9.11~winehq1-1 (diff) [ 47.1kB] Fetched 3B in 0s (4B/s) Skipping unpack of already unpacked source in wine-0.9.11~winehq1 dpkg-buildpackage: source package is wine dpkg-buildpackage: source version is 0.9.11~winehq1-1 dpkg-buildpackage: source changed by Scott Ritchie <scott(a)open-vote.org> dpkg-buildpackage: host architecture amd64 debian/rules clean dh_testdir dh_testroot rm -f build-stamp # Add here commands to clean up after the build process. /usr/bin/make distclean make[1]: Entering directory `/home/user1/apt_sources/wine-0.9.11~winehq1' make[1]: *** No rule to make target `distclean'. Stop. make[1]: Leaving directory `/home/user1/apt_sources/wine-0.9.11~winehq1' make: [clean] Error 2 (ignored) #-/usr/bin/make -C documentation clean dh_clean debian/rules build dh_testdir # Add here commands to configure the package. CFLAGS="-Wall -g -O2" ./configure --host=x86_64-linux-gnu --build=x86_64-linux-gnu --prefix=/usr --mandir=\${prefix}/share/man --infodir=\${prefix}/share/info checking build system type... x86_64-pc-linux-gnu checking host system type... x86_64-pc-linux-gnu checking whether make sets $(MAKE)... yes checking for x86_64-linux-gnu-gcc... gcc -m32 checking for C compiler default output file name... configure: error: C compiler cannot create executables See `config.log' for more details. make: *** [config.status] Error 77 Build command 'cd wine-0.9.11~winehq1 && dpkg-buildpackage -b -uc' failed. E: Child process failed Please advise. **

2 1

GPG test - ignore
by Stefan Dösinger April 15, 2006

April 15, 2006

I wonder why the gpg signatures of the mails I send to wine-patches are wrong when I receive them back, but not wine-devel. Trying to send a mail with a small attachment to here to test. I appologise for the spam. Stefan

1 0

Re: advpack: Perform install initialization in install_init
by James Hawkins April 15, 2006

April 15, 2006

On 4/13/06, James Hawkins <truiken(a)gmail.com> wrote: > Hi, > > Changelog: > * Add a structure to track the data for an install instance. > * Add install_init/release to perform install initialization. > Go ahead and trash all these advpack patches. There are a couple bugs, and some things I'm changing. I'll send the new patches next week. -- James Hawkins

1 0

re: Summer of Code Project: DirectPlay
by Dan Kegel April 15, 2006

April 15, 2006

Alex wrote: > This is a proposal for a Google Summer of Code Project. > > Improve Wine's DirectPlay implementation so that it can at least run a > simple demo app, and ideally end up enabling network play for a free > game demo. > A list of such demos is available here > http://wiki.winehq.org/DirectPlayGames I think Alex is suggesting a possible project idea -- he's not proposing to do it himself. - Dan -- Wine for Windows ISVs: http://kegel.com/wine/isv

1 0

Re: Summer of Code Project: DirectPlay
by James Hawkins April 15, 2006

April 15, 2006

On 4/15/06, "Alexander N. Sørnes" <alex(a)thehandofagony.com> wrote: > This is a proposal for a Google Summer of Code Project. > > Improve Wine's DirectPlay implementation so that it can at least run a > simple demo app, and ideally end up enabling network play for a free > game demo. > A list of such demos is available here > http://wiki.winehq.org/DirectPlayGames > Do you mean this is a proposal for you to work on, or just throwing it out there as a suggestion? -- James Hawkins

1 0

Starcraft vs. fullscreen
by Molle Bestefich April 15, 2006

April 15, 2006

These lines could explain why Starcraft is unable to switch to fullscreen: err:xrandr:X11DRV_XRandR_GetCurrentMode In unknown mode, returning default fixme:xrandr:X11DRV_XRandR_SetCurrentMode Cannot change screen BPP from 16 to 8 fixme:xrandr:X11DRV_XRandR_SetCurrentMode Cannot change screen BPP from 16 to 8 I seem to remember that BPP cannot be changed _at all_ with XRandR, is this correct? If it is, then the obvious thing to do is to leave BPP alone and just change to the correct resolution. What needs fixing here?

10 21

New SetThreadPriority patch
by Mike Hearn April 15, 2006

April 15, 2006

This one uses POSIX capabilities to drop all root privs except for CAP_SYS_NICE, therefore, this is reasonably secure. There is one catch. For some reason a suid root app cannot read /proc/self/exe so relocatability isn't used, and anyway it'd be insecure even if it could as you could hard link wineserver then trick it into loading a malicious library relative to $ORIGIN. I think I will investigate this a bit more, but perhaps later. For now this is fine for RPMs and packages etc, which install to /usr, as they can simply "chmod +s wineserver" and have apps with solid audio. thanks -mike

1 0

Re: Implement THREAD_PRIORITY_TIME_CRITICAL
by Aneurin Price April 15, 2006

April 15, 2006

Mike Hearn wrote: >> I'm not convinced this is true. At least some (maybe most or all) of >> the games showing this problem work just fine if true OSS (ie. not >> ALSA-emulated OSS) is used as the sound driver. WoW and StarCraft >> spring to mind immediately. Plus apparently they work in Cedega >> without needing to be root, though I have no first-hand experience of >> this. > > Maybe for you, but this problem seems to be related to all kinds of > things ... system speed, kernel scheduler, driver combinations, what > Wine is doing at the time etc. All I know is that this works great for > me, and no, rebuilding my kernel and going back to OSS is not really an > option :) > > Cedega, IIRC, either has some awful hacks or has the wineserver acquire > the privs, I don't remember which .... it'd be nice to find out. > > IMHO the best way would be for wineserver to be suid root and then drop > privs. It works everywhere and is simple. I really ought to remember to use reply-all for wine-devel. I've quoted my first post since I meant to send it to the list. Anyway, surely the `best' way would be for the kernel to support user-level `real-time' priorities like the ck kernels. Anybody know why they don't like the idea of that kind of thing? More on topic, does this simply change Wine's priority or does it act on a per-thread level? Most of the issues I've seen have been caused by the audio thread being starved by the others, and is often semi-solved by running Wine at nice 19, which seems counter-intuitive but appears to get around some sheduling problems (priority inversions spring to mind). This has the side effect of course that you have to make sure no other process is going to steal the cpu time from under you. Am I talking about the same issue as everyone else or have I got the wrong end of the stick?

10 36

Knights and merchants (valgrind)
by Joris Huizer April 15, 2006

April 15, 2006

Hello, As suggested earlier on the list, I ran this game with valgrind (at last I tried again with svn, which works better than trying to patch 3.1.1) There are a few problems: valgrind comes with loads of errors, but only indicates which file (*.so) the errors come from (so I need to compile wine with gdb debugger support, without optimisations, I guess?) More serious, it seems wine uses some feature currently not supported by valgrind: ----- x86toIR: unimplemented feature vex: the `impossible' happened: x86 segment override (SEG=CS) prefix vex storage: T total 1908936388 bytes allocated valgrind: the 'impossible' happened: LibVEX called failure_exit(). ==14207== at 0xB00170DB: (within /usr/local/lib/valgrind/x86-linux/memcheck) sched status: running_tid=1 Thread 1: status = VgTs_Runnable ==14207== at 0x480D03: ??? Thread 3: status = VgTs_WaitSys ==14207== at 0x200546A1: (within /lib/tls/libpthread-0.60.so) ==14207== by 0x202E81B7: (within /usr/local/lib/wine/ntdll.dll.so) Thread 4: status = VgTs_WaitSys ==14207== at 0x20129523: poll (in /lib/tls/libc-2.3.2.so) ==14207== by 0x23F252EF: (within /usr/local/lib/wine/wineoss.drv.so) Thread 5: status = VgTs_Runnable ==14207== at 0x2011D347: sched_yield (in /lib/tls/libc-2.3.2.so) ==14207== by 0x23F252EF: (within /usr/local/lib/wine/wineoss.drv.so) Thread 6: status = VgTs_Runnable ==14207== at 0x219840BE: (within /usr/local/lib/wine/winex11.drv.so) ==14207== by 0x23F252EF: (within /usr/local/lib/wine/wineoss.drv.so) Note: see also the FAQ.txt in the source distribution. It contains workarounds to several common problems. If that doesn't help, please report this bug to: www.valgrind.org In the bug report, send all the above text, the valgrind version, and what Linux distro you are using. Thanks. ----- is there a patch for this one? at least I will recompile wine with gdb debugger support and look at the errors valgrind comes with regards, Joris

1 0

Re: Dmitry Timoshkov : winspool: Add a test for GetPrinterDriver, make it pass under Wine.
by Detlef Riekenberg April 15, 2006

April 15, 2006

Am Freitag, den 14.04.2006, 12:46 -0500 schrieb Alexandre Julliard: > Author: Dmitry Timoshkov <dmitry(a)codeweavers.com> > Date: Fri Apr 14 22:34:57 2006 +0900 > > winspool: Add a test for GetPrinterDriver, make it pass under Wine. Thanks for working on winspool.drv, but i'm Sorry that I had not enough Time to look at the Patch before the commit. - Version 0x0400 was correct for win9x-Drivers (4.00). Version 3 is for Usermode-Drivers on w2k and above. (Version 2 is Kernelmode-Driver for NT4.0 and w2k) (Version 1 is NT3.5x and Version 0 is NT 3.1) Our Registry-Location is for win9x only (OpenDriverReg) - One Terminating Zero is is not enough for REG_MULTI_SZ http://bugs.winehq.org/show_bug.cgi?id=3589 - Why was (PDRIVER_INFO_3W) changed to (PDRIVER_INFO_2W)? (PDRIVER_INFO_3W) is an extended Version of (PDRIVER_INFO_2W) - Not all Levels are supported on all Windows-Versions http://msdn.microsoft.com/library/en-us/gdi/prntspol_7xo2.asp - The tests are not alphabetical sorted after this commit - Why was SetLastError(0xdeadbeef) used, when there is SetLastError(MAGIC_DEAD) in the other tests - 11 Tests are now failing on win95, 11 Tests on win98, 9 Tests on winme and 8 Tests on NT3.51. For NT4.0 (level 4, 5 and 6 as well as the Version-Test will fail) and w2k (version will fail, when an NT4.0-Driver is used), i can not test the exact number of failures atm. What's the preferred way to fix this? My Idea: 1. Move the Test without any Changes 2. Fix the Tests 3. Fix the Implementation I know, that you want a fast fix for your commercial Application and winspool.drv is very ugly. I'm on the way to fix winspool.drv, but my Direction is going from the Bottom-Layer to the High-Level API. "struct printerenv_t" was created (as requested by Alexandre) to handle the the different Printing-Environments for win9x-Drivers (Windows 4.0) and NT-Drivers (Windows NT x86). My fix for OpenDriverReg already need an Extension here, but DeleteMonitor comes first. You can Read more Infos here: http://wiki.winehq.org/Printing -- By By ... ... Detlef

2 1

re: user: Fix disabled menu item behavior
by Dan Kegel April 15, 2006

April 15, 2006

Tommy Kho wrote: >This patch changes the behavior of selecting greyed out menu items via >accelerator key. Wine currently keeps the menu open; Windows closes the menu. > ... > else >+ { > pmt->hCurrentMenu = MENU_ShowSubPopup(pmt->hOwnerWnd, hMenu, TRUE, wFlags); >+ return -1; >+ } > >- return -1; >+ return -2; > } Err, isn't the return -2 at the bottom dead code now? That seems wrong... - Dan -- Wine for Windows ISVs: http://kegel.com/wine/isv

2 1

Summer of Code strategy
by Dan Kegel April 15, 2006

April 15, 2006

Hi everybody! There are two weeks before students can submit Summer of Code proposals. How can we use that time to maximize the number of successful projects this year? The students who already have their own good ideas (whoever they are) will probably succeed; we probably need to provide a set of prefab projects of sorts to give guidance to students who are new to Wine and haven't any idea what to work on. Ideally, each such prefab project would have an active developer (or a close approximation) as project leader and mentor. Are any developers willing and able to come up with a project idea and help mentor/supervise a student to work on it? - Dan -- Wine for Windows ISVs: http://kegel.com/wine/isv

1 0

SETI@Home
by Tom Spear April 14, 2006

April 14, 2006

This is a lil bit off topic, but I thought I would bring it up, since we have users that are trying to run windows BOINC under wine. The wine project has a SETI(a)Home team (has had for some time), so if you run the windows or linux BOINC/SETI(a)Home client, you can join our team and apply your credits towards the project's total. Just go to http://setiathome.berkeley.edu/team_display.php?teamid=38091 and click the Join link. Tom

5 5

Summer of Code 2006 is on!
by Dan Kegel April 14, 2006

April 14, 2006

I've updated http://wiki.winehq.org/SummerOfCode, but it's a bit short on project ideas. (Somewhat on purpose -- last year, people just copied and pasted our detailed project ideas. This year, lets give fewer details and more ideas... if somebody's really interested, they'll ask for details on wine-devel, and that process will generate fewer boilerplate applications.) - Dan -- Wine for Windows ISVs: http://kegel.com/wine/isv

1 0

Re: Change threading defaults on FreeBSD
by Alexandre Julliard April 14, 2006

April 14, 2006

Gerald Pfeifer <gerald(a)pfeifer.com> writes: > ChangeLog: > On FreeBSD, build wine-pthread in addition to wine-kthread and make > the former the default. If you build both, then you need a way to decide which one will be used. That's not the sort of detail we want to have users worry about. -- Alexandre Julliard julliard(a)winehq.org

2 2

Fwd: [Bug 4051] winecfg crashes when clicking on Audio tab due to arts bugs
by Tom Spear April 14, 2006

April 14, 2006

---------- Forwarded message ---------- From: Wine Bugs <wine-bugs(a)winehq.org> Date: Apr 14, 2006 7:17 AM Subject: [Bug 4051] winecfg crashes when clicking on Audio tab due to arts bugs To: wine-bugs(a)winehq.org http://bugs.winehq.org/show_bug.cgi?id=4051 ------- Additional Comments From alexander.heinlein(a)web.de 2006-14-04 07:17 ------- Hi. I encounter the same issue. wine wants to access /tmp/ksocket-$USER, but i do not have KDE installed (only KDM) and do not use arts (arts isn't even installed, only some library headers which i can't uninstall because of SDL) or any other soundserver, but hardwaremixing. So i have to manually create the directory /tmp/ksocket-$USER to let the audio tab in winecfg not crash. I tested this with wine 0.9.10 and 0.9.11. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are the assignee for the bug, or are watching the assignee. --------------------------------- Notes from Tom: This should be a fairly trivial fix, all we need to do is check for the existence of the socket file before trying to connect to it, and if it is not there, then skip the arts probe.. Could someone whip up a patch for this? Tom

1 0

DllOverrides
by Rich Gilson April 14, 2006

April 14, 2006

I'm looking at what is done in Winetools (no, this is not another WineTools discussion ;-) ) as far as the DllOverrides and it has generated a few questions. Here are a couple of lines from the wt-config.reg file: "*"="native, builtin" "*autorun.exe"="native,builtin" "msi"="native, builtin" "odbc32"="native, builtin" Can somebody explain the lines with an *? Does the "*" mean that if the driver, dll, exe in question is not listed in the DllOverrides section, that it defaults to using the native version of the file and, if not available, then tries to find and run a builtin version? In the case of the "*autorun.exe", does it literally me that any .exe file with a name ending in autorun would use a native version first, like a normal wildcard would be used? That wouldn't make sense to me because there is no file that matches that description in my .wine directory,. In fact, the only copy of autorun that I can find on my system is the autorun.exe.so from cxoffice. I would appreciate all the enlightenment I can get on this subject. Thanks. -- -- In a world without fences, who needs Gates?

3 3

Hidden files
by Rich Gilson April 14, 2006

April 14, 2006

Here's another question that has been spawned by looking through the Winetools code. It touches two files: ~/.wine/.no_debug_window_flag ~/.wine/.no_prelaunch_window_flag Have these been depricated or completely removed, or are they still something that would be used by Wine? If they are still used, is there a list of all such hidden files and what they do? -- -- In a world without fences, who needs Gates?

2 2

Cups printer detection in pre-built packages
by J.R. Oldroyd April 14, 2006

April 14, 2006

I've been discussing this with the maintainer of the FreeBSD wine port. Right now, detection of libcups.so is done by the configure script at compile time. This means that pre-built packages must either pull in cups as a port dependency in order to always enable printer support, or they must do without printer support. Pulling in cups as a dependency is expensive for those that don't want it, and so cups is disabled for everyone using the FreeBSD binary packages. You need to compile the port if you want cups support. Wouldn't a better detection method be to check at runtime for the presence of a libcups.so (by means of a dlopen()) and enable printers if it's found, or disable them if it's not found? The significance of this would be that a cups dependency is required to build the binary package, but the dependency is optional at runtime. And this means that printers would always work, even from a binary package install. -jr

2 1

[gdi] bug reading/writing metafiles on sparc
by Eric Frias April 14, 2006

April 14, 2006

Due to little-endian/big-endian differences, winelib on Sparc can't read windows metafiles generated on a PC (or on Wine on x86), and conversely PCs can't read metafiles created on a Sparc. It looks like there are a few ways this could be fixed, and I wanted to make sure we don't go about it the wrong way (and that nobody else is working on it). I think the right thing to do is make the on-disk and in-memory formats always be little-endian. We would only convert the data to big-endian format when it is generated or used. GetMetaFileBitsEx and SetMetaFileBitsEx would return & take the raw data in little-endian format. The implementation we're considering involves: - in PlayMetaFileRecord(), convert endianness of paramters as they're used: LineTo(hdc, (SHORT)mr->rdParm[1], (SHORT)mr->rdParm[0]); becomes LineTo(hdc, (SHORT)convert_word_from_little_endian(mr->rdParm[1]), (SHORT)convert_word_from_little_endian(mr->rdParm[0])); - in MFDRV_MetaParam2(), convert endianness *(mr->rdParm) = param2; becomes *(mr->rdParm) = convert_word_to_little_endian(param2); - some similar conversions necessary for anything that reads/writes the header - anything that calls MFDRV_WriteRecord() would need special handling. These seem to be the functions that embed bitmaps or text in the METARECORD, so we'll have to be selective about what gets converted and what gets written as-is. Is this the right way to go about fixing the problem? Anything we should look out for? Thanks, Eric

2 5

Re: Eliminate dependency on Fontforge
by gslink April 14, 2006

April 14, 2006

As I see it, Fontforge was never the problem. The problem was that it was necessary to read the code to find out that there even was such a dependency. This was especially true of the Red Hat binaries. There was no notice or documentation that they would use Fontforge. I didn't find out until I started playing with Fontforge and had Wine crash as a result. It doesn't make any difference what the Wine dependencies are if: 1. They are documented. 2. I can find the dependent programs.

1 0