The original FreshBSD was written 5 years ago, targeting Ruby on Rails and MySQL, and being driven by parsing CVS commit mailing lists. It's messy and barely maintained, and now no longer works for either FreeBSD or DragonFlyBSD because they've migrated to SVN and git respectively, and I'm too lazy to rewrite the mail parsers.
FreshBSD v2 was written 3 years ago, targeting Ramaze and Apache solr, and being driven off local SVN, CVS and git mirrors. It's also somewhat messy, and now no longer works because of a botched attempt to upgrade solr and its associated thousand line XML configs without taking adequate backups.
FreshBSD v3 is currently in development, targeting Padrino and elasticsearch, and driven off tidied up versions of the indexers I wrote for v2.
Development is currently at the prototype stage, but progressing swiftly and showing excellent promise — elasticsearch has shown itself to be pleasant to work with, very flexible and significantly faster than solr, and Padrino gives me a much nicer framework to develop in without the overcomplexity and dubious opinions of Rails.
Hopefully I should have something I'm happy to replace v1 and v2 with in the next week or two.
UPDATE 2011-08-31: beta.freshbsd.org is now running v3 code. It should really be alpha, but meh. Also note it's now indexing FreeNAS, pfsense and PC-BSD, though all imports are ran by hand right now. Getting there…