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

Request for comments on listview patch which was rejected.
by Peter Dons Tychsen 04 Mar '08

04 Mar '08

Hello Devs. I sent in this listview patch some while ago, but it was silently rejected (like most of my patches :-)). Before i re-send it, could someone take a peak at it and tell me what is wrong? It looks quite sober to me... /p

2 1

Bugzilla's down again
by Austin English 03 Mar '08

03 Mar '08

Someone with access care to reset mysql? -Austin

3 3

Building dlls/oleaut32/tests/olepicture.c under Cygwin?
by Dan Kegel 03 Mar '08

03 Mar '08

I tried building dlls/oleaut32/tests/olepicture.c on Windows using cygwin, but ran into a couple small problems. The main problem is: /cygdrive/c/DOCUME~1/Liz/LOCALS~1/Temp/ccXL7SF3.o:olepicture.c:(.text+0x18e5): undefined reference to `_IPictureDisp_Invoke' That sounds fishy - is that test using a wine-specific shortcut? If so, what's the kosher windows way to do what it's doing? Here's the procedure I followed: 1. Download and run the installer from http://cygwin.com Use it to install gcc and git 2. Open a cygwin window 3. Grab a copy of the Wine sources using git as normal 4. cd wine/dlls/oleaut32/tests 5. to avoid conflicts between the win32 headers from wine and from cygwin, just use the cygwin ones. But we do have to copy one header from wine: mkdir -p include/wine cp ../../../include/wine/test.h include/wine 6. Try compiling and linking with $ gcc -DSTANDALONE -Iinclude olepicture.c -mwindows -loleaut32 -lole32 -luuid This fails with all sorts of errors about LOGFONTA. That's because the include lines in olepicture.c are in the wrong order! Edit olepicture.c and change the lines #include <winuser.h> #include <wingdi.h> to read #include <wingdi.h> #include <winuser.h> 7. Try compiling again. Now you fail with a handful of errors like /cygdrive/c/DOCUME~1/Liz/LOCALS~1/Temp/ccXL7SF3.o:olepicture.c:(.text+0x18e5): u ndefined reference to `_IPictureDisp_Invoke' Seems like a portability problem in tests/olepicture.c.

3 5

Next big app(s) to try?
by Dan Kegel 03 Mar '08

03 Mar '08

Now that Adobe Photoshop CS2 is running kind of well (there are bugs that need fixing, but I'm trying to look ahead a bit), what big app(s) are worth focusing on next? Using http://wiki.winehq.org/LinuxApplicatonRequestSurvey, skipping Photoshop (which we should continue to fix up) and the ones that are either too hard, have native versions, or are probably something Codeweavers might already be doing, and adding in good old Framemaker (so close to working we might want to go for it just for the goodwill from all those tech writers), I think that leaves about seven big ticket apps, in rough order of popularity: http://wiki.winehq.org/AdobeDreamweaver iTunes http://wiki.winehq.org/AdobeIllustrator http://wiki.winehq.org/AdobeFlash http://wiki.winehq.org/AdobePremiere http://wiki.winehq.org/AdobeAcrobatPro http://wiki.winehq.org/AdobeFramemaker I've created wiki pages with trial download URLs for all but iTunes, and done a bunch of triage on them. Most of these could all use more triage and bugs filed. At some point the dust might settle, and it might become clearer which one(s) deserve to be focused on. I think Maarten is going to try iTunes, which would be great. - Dan

9 10

Re: winex11.drv: Add DefaultRefreshRates option to specify refresh ratewhen application does not
by Dmitry Timoshkov 03 Mar '08

03 Mar '08

"Denver Gingerich" <denver(a)ossguy.com> wrote: > This patch allows the user to specify a list of default refresh > rates to use at particular resolutions to override the default > behavior that chooses the refresh rate based on whichever mode > is first in the list for a given resolution. The list is specified > by populating the X11 Driver/DefaultRefreshRates registry key with > a semicolon-delimited list of resolution, refresh rate pairs. What's really needs to be done is writing the tests to investigate what Windows does in that case, and make the tests pass in Wine. -- Dmitry.

2 3

Re: [PATCH 1/4] rpcss: Implement endpoint-mapper server using DCE/RPC.
by Alexandre Julliard 03 Mar '08

03 Mar '08

Robert Shearman <rob(a)codeweavers.com> writes: > --- > include/Makefile.in | 2 + > include/sharedidl/dcetypes.idl | 88 ++++++++++++ > include/sharedidl/epm.idl | 113 ++++++++++++++++ > programs/rpcss/Makefile.in | 3 + > programs/rpcss/epm.idl | 19 +++ > programs/rpcss/epmp.c | 291 > ++++++++++++++++++++++++++++++++++++++++ > programs/rpcss/rpcss_main.c | 54 +++++--- > 7 files changed, 550 insertions(+), 20 deletions(-) > create mode 100644 include/sharedidl/dcetypes.idl > create mode 100644 include/sharedidl/epm.idl > create mode 100644 programs/rpcss/epm.idl > create mode 100644 programs/rpcss/epmp.c That sharedidl stuff is ugly, we really have to find a way to make this work with the idl files in include/wine. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [1/3] shlwapi/tests: Move the function pointers to the string test functions where they are initialized and used.
by Alexandre Julliard 03 Mar '08

03 Mar '08

Francois Gouget <fgouget(a)free.fr> writes: > It makes no sense to declare the function pointers globally, but then > only initialiaze them in one function or another. This is a recipe for > adding code that calls the function pointer elsewhere... and before it > has been initialized of course. Actually the reason for declaring them globally is that 2.x gcc versions sometimes crash on local WINAPI function pointers. -- Alexandre Julliard julliard(a)winehq.org

2 1

Re: [PATCH 1/6] OLE: Remove tabs from olepicture.c test file.
by Alexandre Julliard 03 Mar '08

03 Mar '08

Please don't do that. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [PATCH 1/1] GDI+: Implement GdipGetImageEncoders.
by Alexandre Julliard 03 Mar '08

03 Mar '08

Nathan Beckmann <nathan.beckmann(a)gmail.com> writes: > + > +/************************************************************************* > + * Encoders - > + * Implemented as dummy objects that let the user know which files are > + * supported. Actual encoding is done by OLE. > + * > + * This code is based heavily on libgdiplus (Mono): > + * > [...] What's the point of including such a large copyright notice if you are not actually copying any code? You are just defining a couple of constants, you don't need to copy that from Mono. -- Alexandre Julliard julliard(a)winehq.org

1 0

mountmgr.sys and ioctl's
by Maarten Lankhorst 03 Mar '08

03 Mar '08

Hi Alexandre, In irc we discussed ioctl's with devices, and how ioctl's go to ntdll/cdrom.c if the drive is fd backed, and to mountmgr.sys if it isn't. For itunes it needs raw disk access, but also an ioctl from mountmgr. What if we always map access to \\.\X: is always mapped to mountmgr, and let mountmanager open \\.\X:: that would be mapped to /dosdevices/x::, any ioctl, and read/write stuff that mountmgr can't handle can then be forwarded to the raw device, if available. Do you think that would be an appropiate solution? Cheers, Maarten.

2 1