On Mon, 1 Jun 2009, Richard W.M. Jones wrote:
Large data sets derived from analysing the code in Fedora packages... I was hoping to save the announcement til later. For the purposes of this, big data files, derived from Fedora content so under a free license of some sort, probably up to 30GB in size. They would be regenerated once a month.
30GB once a month shouldn't be too much trouble, the biggest problems will be:
1. transferring the data in a timely-ish manner2. keeping lots of those datasets around. Hopefully no more than 2 b/c 60GB is a big chunk of space
-sv