search @ delicious and the bbc
I just noticed that del.icio.us now has full, fast search across all content (not just your own bookmarks). This is something that Dan’s unalog has had on delicious for a while (apart from the delightful content). Dan uses pylucene as his search engine, which still has some interesting features. It’s pretty wild being able to search across all the delicious content, given their volume.
When delicious was really ramping up I saw the occasional mason error page, so I know that they are (or were) using Perl. This makes me really curious to know what search technology they are using…but I couldn’t find any details in the announcement.
Likewise, the news about the BBC Programme Catalogue being built with RubyOnRails. I’ve really come to appreciate Lucene and PyLucene and am in search of similar search tools for Ruby. I’ve got an email out to Matt Biddulph to see if he can provide any details about the BBC effort.
November 10th, 2005 at 5:31 am
How about Ferret:
http://ferret.davebalmain.com/trac/
which seems to be a Ruby port of Lucene.
I’ve been lurking on the Ruby Weekly news page (and various others) trying to figure out if I want to invest the time learning. So far, a thumbs up.
November 10th, 2005 at 8:40 am
I’ve been looking for a similar search tool for Ruby. Have you seen ferret? It is a port of Lucene to Ruby. I guess the performance isn’t great right now, and it’s not terribly finished, but it should be very useful as development continues.
http://ferret.davebalmain.com/trac/
I’ve also been meaning to play around with ruby/odeum. I understand that performance is pretty good, but it’s only for indexing; it doesn’t do any of the lexical parsing that I understand Lucene performs.
November 10th, 2005 at 9:12 am
Many thanks Eric and jonvw — I’ll take a look at Ferret. I missed you guys at the ruby meetup on Monday. Hopefully we can get together next time.