On Dec 20, 2007 9:10 AM, Luca de Alfaro luca@soe.ucsc.edu wrote:
That's true. We had to truncate histories to make everything fit into a server. We are gaining experience in how to deal with Wikipedia information (terabytes of it), and we may be able to give a better demo in some time, with full histories, but.... we need to buy some storage first! :-)
Could you possibly take a random sample of 2% of articles and examine the full histories of those? 40,000 articles is more than enough for a dem, and we can rig the sample to include some articles of interest if needed.
Akash