Re-encoding old videos from Flash to another format seems like a pretty basic hadoop-style problem that you could easily implement on EC2. Of course this isn't

Re-encoding old videos from Flash to another format seems like a pretty basic hadoop-style problem that you could easily implement on EC2. Of course this isn't free, but I remember reading something about NYT converting 100 years of archives to PDFs for a rather trivial amount of money (and time) by doing it this way.