Are there any sources of real raw data I can play with for free ?
I don't care what the data is for, as long as there is a LOT of it.
The more (in both volume and broadness of scope), the better. :)
//Edit to add -- Text data, binary data, images, doesn't matter what it is. A note about the type is nice though.
The Federal Aviation Authority publishes its Aircraft Registration database online. This contains hundredds of thousands of records. Two caveats. The Addresses are American (which probably won't bother you) and the Onwers mix indiduals with companies (which might). Still it's a decent source of data with built-in application logic.
Download it from the FAA website.
Take a look at Stack Overflow data dump.
Also I've seen some people extracting massive amounts of data from Twitter via its API. Can't find the link now though...
Astronomical star catalogs are pretty huge; they contain information on millions of stars, usually in text table format. Most of them are freely available.