Hi,
I'm interested in working with you as a student to create a GStreamer plugin for Wine.
First, a little background about me. I'm a 2nd year student from Bucharest, Romania; my major is Systems Engineering. Last summer I spent my time developing kmemtrace (a memory tracer for the Linux kernel) under the mentorship of Pekka Enberg penberg@cs.helsinki.fi from Linux Foundation, that is GSoC 2008. So this would be my second year in GSoC. And I do tend to stick around after the SoC is over, since it's a great opportunity to get comfortable with a code base. You can check out my coding style and performance by looking at these Git repos:
Ingo Molnar's -tip tree (my current "upstream"): http://git.kernel.org/?p=linux%2Fkernel%2Fgit%2Ftip%2Flinux-2.6-tip.git&... My repo holding kmemtrace's userspace app: http://repo.or.cz/w/kmemtrace-user.git
However, I'm aware of the differences in coding style between Wine and the Linux kernel. Regarding my skills, I'm quite proficient at using Git (managing my own tree and sending patches), coding in C and using autoconf. I also know some very basic concepts about DirectX, OpenGL and Wine's architecture.
Also, this isn't my first contact with Wine's source code. Even though I haven't sent any patches to the mailing-list, I've spent some time reading the source code (chasing performance issues in Direct3D, but to no avail). You might have also seen me on IRC, bugging people about those issues. :-)
Now getting to the point... I find the GStreamer plugin a very good idea. Glancing over GStreamer docs, I see it's very modular and could solve more issues than those already mentioned on the GSoC ideas page. I'll focus on what hasn't already been told.
Firstly, it could also improve GDI+'s handling of images, as GStreamer already handles some image formats. Secondly, I've seen a patch implementing texture handling in d3dx being rejected because there is no general image library support in Wine and FreeImage wasn't a comfortable dependency: http://archives.free.net.ph/message/20080828.215032.87fd785e.en.html
So this work could also have a positive impact on the proposal regarding the implementation of d3d9_xx DLLs. Bottom line is this won't "just play movies for DirectShow/Quartz", but it has a much greater potential.
I admit, my main interest in developing Wine is improving a few games (e.g. Oblivion). I find this GSoC idea quite achievable for my skills and positive towards this goal. I also hope this experience would make a good introduction to implementing better DirectX support on Wine, generally speaking.
Please let me know what you think and whether I've missed something. And please keep me Cc-ed if more people pitch into the discussion, as I'm not subscribed to the mailing-list.
Cheers, Eduard