Google Announces Summer of Code 2008 Projects
An anonymous reader writes "Google announced today it had accepted 1,125 students to work on 175 Free and Open Source Projects this summer. This represents an increase of almost 25% over last year. Nearly 7,100 applications were received. For those who weren't accepted, there is an offer to send Google Swag to any student who completes their project anyway."
...that I hate Google for their immense privacy violations, and yet can't help but get excited when the Summer of Code comes up?
It'll be interesting to see how they are integrated and how big a change some of those items become at the other end of SoC.
Title Generic frame-level multithreading support
Student Alexander James Lloyd Strange
Mentor kristian Jerpetjoen
Abstract
FFmpeg, while equalling or surpassing the speed of nearly all other codec implementations on a single CPU core, currently only has limited and specific support for multithreading. I will implement a frame-level multithreading system, which can efficiently speed up all uses of libavcodec. This will be based on the successful implementation in the x264 encoder[1], extended to support decoding and whatever synchronization will be required. [1] http://akuvian.org/src/x264/sliceless_threads.txt, http://www.techarp.com/showarticle.aspx?artno=442&pgno=0
Mono's C# is extremely slow compared even to OpenJDK Java, saying nothing of finely optimised C. It would be very nice if modern optimisations could be brought over from OpenJDK to Mono, but I doubt that will happen anytime soon.
Sam ty sig.
I attribute my proposal being accepted to the fact that I'm implementing a real-time web version of a game that makes my mentor (and probably other proposal-choosers for Portland State University) feel quite nostalgic.
When I called him to initially discuss the idea, he actually cut me off mid-sentence and said--with Renee-Zellweger-like tearful joy in his voice--"You had me at 'Nomic'."