March 2008 - Wine-devel - list.winehq.org

Re: [Wine] Re: How to know which env vars a software wants
by Dan Kegel 17 Mar '08

17 Mar '08

vitamin <wineforum-user(a)winehq.org> wrote: > > > A simple touch ~/c:/windows/system32/drivers/etc/{services,host} > > So perhaps this will help other Informix-based apps? > > Should probably add these files to Wine tree. "hosts" shouldn't need > anything more then standard 127.0.0.1 localhost > And services should be empty. Yep. Filed as http://bugs.winehq.org/show_bug.cgi?id=12076

1 0

Updated 1.0.0 release criteria; draft 1.0.0 release plan
by Dan Kegel 17 Mar '08

17 Mar '08

I've updated http://wiki.winehq.org/WineReleaseCriteria to be a bit more final. I've also written a draft release plan; see http://wiki.winehq.org/WineReleasePlan Comments?

5 9

Re: [Tools 7/7] winetest: Add some heuristics to detect missing dlls in version 3 reports.
by Paul Vriens 17 Mar '08

17 Mar '08

Hi, I saw that 'gather' is not creating the index.html file at test.winehq.org anymore. I had a quick look at test.winehq.org and something seem to have go wrong with the patching of 'gather': pvriens(a)wine:/home/winehq/opt/winetest$ perl gather Use of bare << to mean <<"" is deprecated at gather line 142. Use of bare << to mean <<"" is deprecated at gather line 171. Global symbol "$testref" requires explicit package name at gather line 142. Global symbol "$extrema" requires explicit package name at gather line 142. Global symbol "$extrema" requires explicit package name at gather line 142. Global symbol "$extrema" requires explicit package name at gather line 142. Global symbol "$extrema" requires explicit package name at gather line 142. Global symbol "$extrema" requires explicit package name at gather line 142. syntax error at gather line 176, near "HEAD:" (Might be a runaway multi-line << string starting on line 171) Execution of gather aborted due to compilation errors. -- Cheers, Paul.

3 2

Is the cxtest.org site down?
by Maarten Lankhorst 17 Mar '08

17 Mar '08

Hi, I was trying to look at www.cxtest.org but the site seems not to work. Can this be fixed? Cheers, Maarten.

2 1

Improving the Wine Users Guide
by Lei Zhang 17 Mar '08

17 Mar '08

Hi, For our 1.0 release, we should have a polished users guide to help people to learn to use Wine properly. I did a quick read of the Wine Users Guide [1] and here are my comments: * Near the top - mention virus / security issues, warn against running Wine as root. * Remove references to tools/wineinstall - does anyone use this anymore? * Section 2.3.1 - mention ./configure --verbose * Section 3.1.2.1 - are there dlls besides kernel32, gdi32, user32, and ntdll that we should never use the native one? Add warning about downloading dlls from the web. * Section 3.1.3 - the default virtual window resolution is 800x600 now. Add documentation for the dpi setting. Remove references to "desktop double buffering" - option no longer exists. * Section 3.1.4 - document the advanced options. * Section 3.1.5 - add a blurb about setting directsound emulation to none when wine complains about it. * Section 4.5 - mention the WINEPREFIX environment variable. * Section 5.2.1 - mention including hardware / drivers info in bug reports - Lei [1] http://www.winehq.org/docs/en/wineusr-guide.html

7 6

wine's very own control panel applets?
by pure_evil＠mail.bg 17 Mar '08

17 Mar '08

Hey folks I was thinking of writing some substitutes/stubs for the windows .CPL files. I migrated to using linux as a desktop fairly recently /used it for servers for a while, though/. Adapting to the linux equivalents of the programs I use, using wine for the rest etc. Whilst "reinventing" the software base,I had an annoyingly familiar *close encounter* with a program which /both windows and wine/ gives out a message basically saying in short: "For this program to run, the decimal should be '.' and the negative sign - '-'." That's cuz the locale /Bulgarian/ uses the ',' as the default decimal /both win and wine/. The solutions, of course, are the 3 obvious ones - copy intl.cpl from a working windows, go "regedit" on the subject, open the regs with a text editor. Both 3 are not very appealing (notably the first one; the 2nd and 3rd are rather annoying /if you haven't tampered with the registry too much/, but I consider the first one a real no-go.). So I decided to write something to replace the functions of intl.cpl, and called it "winelocalesetup". However, this got me thinking further - what if that program instead of popping up that message, decided to execute,say, "control.exe intl.cpl", to save me some of the trouble? And didn't issue any warning, but instead crashed in midst of working with it if the decimal wasn't changed (which would normally be the case. It's an accounting program, so, the decimal is indeed important...) Anywho, I was thinking of creating a bunch of CPLs which mimic their windows counterparts in their names. However, the applet itself wouldn't do much - it would basically contain a somewhat appropriate icon/description, and, when started, launch the real executable, wine*something*.exe. That "wine*something*.exe" will for now basically show a message "Winesomething. Not implemented yet. If you see this message because of a running program, it means that the program is trying to...". With time, the .exe will be transformed into a useful substitute for the corresponding windows function. For example: intl.cpl - contains an icon (preferably a picture of the Earth; was thinking of going NASA on the subject), a description "Wine Locale Settings", and will execute winelocalesetup.exe. Winelocalesetup.exe would show a message: "Winelocalesetup. Not implemented yet. If you see this message, it means that a program is trying to call INTL.CPL, probably in order for you to modify the regional settings." That way, at least, it will be indicated to the user what's going on. The idea for having both the cpls and the exes is, well, the cpls have to be called what they're called for compatibility reasons, and, once in place, need not be modified; the actual tool is contained within the .exe, which can be, well, a lot more flexible to develop than the CPL itself, and, can be replaced at any moment. And, well, although compatible with windows, it wouldn't be a pure repetition. (and no longer an empty control panel. Yay!) BTW I have done quite some work on that winelocalesetup.exe, it reads the appropriate values from the registry, doesn't resemble the original intl.cpl's vision, and, well, is somewhat kinda useful (if only I could figure out how to WRITE the damn values, the keys keep turning back to their old values... oh well, guess I'll have to RTFM some more.) So, any suggestions/comments/ideas?

3 4

re: Wine Security Disclosure
by Dan Kegel 17 Mar '08

17 Mar '08

<tehblunderbuss at gmail.com> wrote: > The recent discussion about Windows viruses working through Wine leads > me to questions about its security. > I've heard that using a separate user is alright, and then it isn't > alright. You're probably referring to http://www.winehq.org/pipermail/wine-devel/2008-March/063452.html I was talking about how to make sure your machine is safe when you *on purpose* load malware. However, it is true that we should tell users that Wine is able to run many Windows malware, and that they should be just as careful running Windows software with Wine as they would running it with Windows. We should also show how to install the standard Linux virus checkers. - Dan

6 10

Locked topics considered harmful
by Dan Kegel 16 Mar '08

16 Mar '08

wine-users is not a forum, it's a mailing list. Topics should not be locked except in extreme circumstances; the bounce messages are annoying, and it's wrong for the forum to not properly show what's going on in the mailing list.

2 1

Re: wined3d: Increment reference count before returning palette
by Stefan Dösinger 16 Mar '08

16 Mar '08

Am Sonntag, 16. März 2008 17:49:38 schrieb Cihan Altinay: > Cheers, > Cihan You'll have to check the code that uses this method. As far as I can see DDrawSurface::GetPalette doesn't release the palette it got from WineD3D, so just changing WineD3D will cause a leak

2 1

Re: d3dx9_36: Implements D3DXGetShaderVersion
by Ivan Gyurdiev 16 Mar '08

16 Mar '08

Luis C. Busquets Pérez wrote: > This functions gets the Version Token of a binary shader. This Token > is the first of the binary stream so it is as easy as picling the > pointed DWORD. This is all public information available to the public. What happens when the shader is NULL? Yours will crash - MSDN says otherwise. Does this work on Pixel and Vertex shaders ? Does it always return back the first dword as is - what happens if an invalid shader version is passed in. > To implement a test would be far more complicated as it would mean to > include a binary shader. Anyway, I have tested the results with Civ4 > (by making a derivation on d3dx9_32 with this function and using the > original d3dx9_36.D3DXGetShaderVersion function) and the results where > the same for all the shaders that where present. I would suggest an automated test across [ vs: 1.0, 1.1, 2.0, 2.1, 3.0 ] , [ ps: 1.0, 1.1, 1.2, 1.3, 1.4, 2.0, 2.1, 3.0 ] , NULL, and an invalid token. You can see some binary shaders in the d3d9 tests. Ivan

2 1