Google ruffled a lot of feathers a few weeks ago when it announced that it would be dropping H.264 support from Chrome. In the short term, it seems like a ridiculous move - one guaranteed to kill HTML5 video, and keep Flash dominant for the foreseeable future - but come 2015, I think this move will make a lot of sense.
Some background on video codecs: A video codec is a method of compressing video. The measure of a codec is how well it compresses, and what sort of tradeoff you have to make between file size and image quality. Any codec can give you perfect video if you give it enough space to work with, but H.264 is one of the best when it comes to giving you high quality video at reasonable bitrates. Other prominent codecs for web streaming include Theora (which is open-source and royalty-free, but doesn't compress as well as H.264), Dirac (an experimental codec developed by the BBC, which has yet to gain any traction), and VP8 (Google's codec, bought from On2, and part of WebM). H.264 is dominant right now because it gives good results, is widely implemented, and because hardware accelerated decoders are available (crucial for mobile devices).
The HTML5 video tag doesn't specify any required video codecs, so sites are responsible for using one that all their users are able to use. Last year there was a huge shootout between proponents of H.264 and Theora. H.264 is technically the better solution, but it's owned by the MPEG-LA, and they intend to start charging royalties for it in 2015. Right now, though, H.264 seems to be winning - it's backed by Microsoft and Apple, and it's actually the only way to get video to play on iOS devices.
This is a problem for Google: YouTube would be hit pretty hard by the license fees, since every video on YouTube is streamed in H.264. Google doesn't plan to take this lying down, though. If Theora isn't up to the job, then Google will buy a codec that is competitive with H.264, and release it royalty-free for anybody to use. If it looks like H.264 is still winning, then Google is willing to drop H.264 from Google Chrome (just over 10% of the browser market) to try to kill its momentum, and replace it with Google's own codec.
MPEG-LA was originally set to start charging royalties for web streaming H.264 in 2010, but they moved it back to 2015 in order to allow H.264 to become more solidly entrenched. People say that Google is being shortsighted by promoting their own codec, since they'll never make any progress when IE9 and Safari (and iPhones!) support H.264 by default. Google doesn't need to win, though - they just need a solid alternative to H.264 to exist by 2015, so that MPEG-LA is in a weaker position when it comes time to work out what YouTube has to pay for using H.264. That's their real game here, and so far it seems like they're going to make it.