Sometimes I think about video compression algorithms. Some ideas seem very cool but you can't really figure out how they work until you try to implement them.

Today I wrote a few lines of code to test one of these ideas (see video below).

You may ask what's the point in writing a new codec today. My response is... NO FUCKING REASON! :)
Well, I'm thinking about a uber-lossy codec, where the glitches themselves are someway artistic. I may even synch the music with video artifacts, for example.

Another goal is to write a video format that can be easily played back on the GPU through WebGL.