I love the concept of duolingo.com: learn another language for free while helping to translate the web. It's brilliant, simple, and ticks all the right boxes for a project like that. I've signed up to learn Spanish, but I haven't used it much yet. The one concern I have is that they say their aim is to translate the whole web into all major languages, but I probably produce more English text in a day than I would get through on duolingo in a week, and that's just me. If not everyone is participating, and they have to aggregate translations to get good results, and furthermore the participants are producing more new native language text than they are translating daily, the project is going to start behind and get further behind as time goes on. Granted, it's better than starting behind and staying still, and if you focus your efforts on major websites like Wikipedia, you probably will get through a reasonable amount quickly, so I think it's definitely worthwhile. If you've ever wanted to learn another language (and you probably have) then this is a good way to go, if only because it's free.
Mokalus of Borg
PS - I wonder whether a similar concept could work for programming languages.
PPS - The task would have to be platform-porting.