I apologize that I didn’t post an update earlier. I didn’t do a post for week 8(from July 6-10), and the next week was only a partial week as I left on vacation on Thursday. I was on vacation from the 16th to the 24th. Then I’ve just completed another week of progress. I’ll just sum up everything I’ve done in one big gigantic post.

What I’ve done since then:

• Fixed video sync issues after pausing and unpausing
• Fixed a bug dealing with overlay and surface releasing and resizing
• Expanded unit tests for all the backends
• Added more commenting, cleaning up code
• Began working on subtitle support but discovered that ffmpeg 0.5 has no demuxers for subtitles so decided to delay this feature for after GSoC
• Worked on various references issues
• Moved audio packet processing to its own thread
• Wrote a more optimized routine for writing data to an SDL_Surface, still slow on large screens
• Added some profiling
• Merged branch with trunk revisions
• Fixed slowdown issues
• Added seeking. No relative seeking yet, it is absolute for now.
• Updated the code to work with older ffmpeg libraries. Only one section is incompatible, and waiting on a response from the libav* mailing thread
• Began working on a command queue to make the module more thread safe.

There are some current issues at the moment.

• If using surfaces, and you resize the window smaller, there ends up being corruption in the bottom of the window. I have no idea why this happens.
• I will be adding relative seeking this next week.
• Found an issue where commands issue from the Python interpreter thread in sequence but with no delay cause a big crash. Thus the necessity of the command queue.
• Code seems a bit slow, and very very rarely, audio and video get desynced.
• No path expansion yet.

And there we go!