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

FEAR Combat
by Jeremy Newman Aug. 25, 2006

Aug. 25, 2006

Tried FEAR Combat on Wine. Looks like it needs a d3dx9_27.dll which we don't provide. Anyone else have any luck with this one? I'm not sure what the status is of all the required dx 9.0c dlls are.

9 15

Write-strings warning: the endgame
by Andrew Talbot Aug. 25, 2006

Aug. 25, 2006

Currently, there are just eight write-strings violations remaining in the codebase: six in mshtml/nsembed.c (lines 271, 353, 356, 367, 432 and 1237), one in mshtml/nsservice.c (line 454) and one in widl/parser.c (line 848). So I would advocate that the -Wwrite-strings switch could now be turned on. I have just submitted two speculative patches to attempt to fix these remaining warnings, but, if they are wrong, I would be grateful if those of you with mshtml expertise or yacc expertise should kindly step in and finish the respective job properly. Thanks, -- Andy.

3 3

Re: Write-strings warning fix
by Dan Hipschman Aug. 25, 2006

Aug. 25, 2006

On Fri, Aug 25, 2006 at 11:08:03AM +0100, Andrew Talbot wrote: > What confused me - especially with my ignorance of yacc - is that I couldn't > find a corresponding xmalloc() that assigns the char * variables that the > above free() frees. make_var() gains storage for a var_t type, but the > free() in get_type() releases the storage of a char *. I couldn't quite > follow what was going on, there. Is it because the string is the first > element in the structure, so the pointer to the string also points to the > start of the var_t type? The strings are allocated in the lexer (parser.l): yylval.str = xstrdup(kw); The patch that was actually committed (using xstrdup) is most consistent.

2 1

Re: getwinegit.sh v0.4 released!
by Michael [Plouj] Ploujnikov Aug. 25, 2006

Aug. 25, 2006

In light of the recent discussion about moving to GIT I would like to let send this email to everyone on wine lists. Since Christian's last update to his getwinegit script I have tried to improve it to be more general and work in my special setup. I sent him this same email but I haven't heard a reply yet. ---------- Forwarded message ---------- From: Michael [Plouj] Ploujnikov <ploujj(a)gmail.com> Date: Jul 26, 2006 10:34 AM Subject: Re: getwinegit.sh v0.4 released! To: Christian Lachner <gladiac.lists(a)gmail.com> I have made some fixes to the script and an example config file: - Removed the check for root user. The script shouldn't require root access, not to compile wine, not even to install it. I think the best way is to just check that the user running the script has write permissions to install, log, and git checkout folders. - changed the CONFIGURE variable to be an array to allow for quoted words passed to ./configure. From my configuration that would be something like this: ./configure --x-libraries=/emul/linux/x86/usr/lib/ LDFLAGS="-L/emul/linux/x86/usr/lib -L/emul/linux/x86/lib" (because I need to compile 32bit wine on a 64bit machine with emulation libraries) - fixed one typo that I noticed - added a definition of MAKE to my config file. Without it DEPEND=$MAKE" depend"; and others don't execute make. The same should also be added to the nocolors.config file: MAKE="make"; Of course, you can ignore the path changes in the config file, unless you want an example of what a non-root user would do. The changes are attached as unified diffs. For some reason google made me reply to you rather than wine-devel and I guess it's up to you to post the changes there anyways ... after a thorough review! Enjoy.

1 0

Re: LOSTWAGES: Replace references to CVS with Git
by Robert Shearman Aug. 25, 2006

Aug. 25, 2006

Mike McCormack wrote: >Index: templates/en/development.template >=================================================================== >RCS file: /home/wine/lostwages/templates/en/development.template,v >retrieving revision 1.6 >diff -u -p -r1.6 development.template >--- templates/en/development.template 28 Oct 2005 16:05:10 -0000 1.6 >+++ templates/en/development.template 18 Aug 2006 01:02:06 -0000 >@@ -6,7 +6,7 @@ > WineHQ hosts a number of services aiding the Wine development process. > > <ul> >- <li><a href="{$root}/site/cvs">Wine CVS tree</a></li> >+ <li><a href="{$root}/site/git">Wine Git tree</a></li> > <li><a href="{$root}/site/contributing">Contributing to Wine</a></li> > <li><a href="http://bugs.winehq.org/">Bug tracking</a></li> > <li><a href="http://appdb.winehq.org/">Application Database</a></li> > I think we should leave the reference to the CVS tree page (but still add a link to the Git tree). Some people find it easier to work with CVS than Git. -- Rob Shearman

12 22

Re: LOSTWAGES: Fix the UI status page html and links
by Dimi Paun Aug. 25, 2006

Aug. 25, 2006

On Fri, 2006-08-25 at 19:30 +0900, Mike McCormack wrote: > > Maybe we should move this to the Wiki too? Maybe. But it changes seldom enough, and I think it will be ugly to get all the formatting right in the Wiki format. -- Dimi Paun <dimi(a)lattica.com> Lattica, Inc.

1 0

Re: msi: Store properties in a hash table, not the database.
by Dan Kegel Aug. 25, 2006

Aug. 25, 2006

On 8/24/06, Mike McCormack <mike(a)codeweavers.com> wrote: > Use string insensitive comparisons for property names. > --- > dlls/msi/package.c | 253 package.c: In function 'MSI_CreatePackage': package.c:426: error: 'PROPERTY_HASH_SIZE' undeclared (first use in this function) package.c:427: error: 'struct tagMSIPACKAGE' has no member named 'props' Did you forget to include your change to msipriv.h?

1 0

Finding a regression
by Doug Laidlaw Aug. 25, 2006

Aug. 25, 2006

I may be on the wrong list. A program I use has shown a backward step in graphics between Wine 0.9.15 and 0.9.16, and I am trying to find the change responsible. I currently have Wine set as at 2006-06-21 16:21:20 CDT, it is identifying as 0.9.16, and the fault is present. The Web site says to get the CVS history from the mailing list archives, but the archive is only up to mid-2005. What is the best way to step backwards? Doug. -- When we want to read of the deeds that are done for love, whither do we turn? To the murder column. - G.B. Shaw.

4 9

Re: msi 5: Add initial implementation of the DirectoryListUp event
by Mike McCormack Aug. 25, 2006

Aug. 25, 2006

James Hawkins wrote: > +UINT msi_dialog_directorylist_up( msi_dialog *dialog ) > +{ > + msi_control *control; > + BOOL indirect; > + LPWSTR prop, path, ptr; > + > + control = msi_dialog_find_control_by_class( dialog, WC_LISTVIEWW ); > + indirect = control->attributes & msidbControlAttributesIndirect; > + prop = msi_dialog_dup_property( dialog, control->property, indirect ); I don't think we should be looking up controls by their Window class. Either look up "DirectoryList" by it's MSI control type, or have the DirectoryList code subscribe to the event somehow. > +UINT msi_dialog_directorylist_up( msi_dialog *dialog ); Please put declarations for functions in a different file in msipriv.h Mike

1 0

Re: msi 4: Add initial implementation of the DirectoryCombo control
by Mike McCormack Aug. 25, 2006

Aug. 25, 2006

James Hawkins wrote: > + indirect = msi_dup_property( dialog->package, control->property ); > + prop = indirect; > + } > + > + val = msi_dup_property( dialog->package, prop ); > + lstrcpynW( path, val, MAX_PATH ); > + > + PathStripPathW( path ); > + PathRemoveBackslashW( path ); Calling msi_dup_property allocates new strings, so you need to remember to free those strings too. Since it's allocated, why not use PathStripPathW and PathRemoveBackslashW on the allocated string, rather than trying to copy it first? Mike

1 0