[Tagdb] Nitins talk and future of tag systems

[Philipp Keller]

> I mentioned Simpy on this list before.
> I'm one of the Lucene developers and co-author of Lucene in Action
> (Erik, the other co-author, is also on this list, I believe), so guess
> what powers Simpy?
Hehe.. can you elaborate a bit on the usage of Lucene in Simpy?
Do you do the tag search on Lucene, but other data searches in
PostgeSQL..? I'm curious.

> Simpy is also not using MySQL.  I'd looove to see your benchmarks run
> on other databases, for instance PostgreSQL, or Oracle.  Any chance you
> can do that?
Hmm.. right now I am still on a clustering algorithm (powered by regular
tips from Scott White.. :-), but when this is finished, I think I could
give it a try. 
But I'm more curious in doing tests with Lucene though.. I don't believe
PostgreSQL will be much faster than MySQL, I think the problem of the
lack of performacne in tag-intersection/union queries lays in the nature
of relative databases..

> Regarding comments about the Popular page (and CSS and such).... I
> can't talk about it yet, but keep an eye on Simpy.  What you addressed
> is precisely what the upcoming release will address.  I better hurry up
> and get to work, now that I said this in public.
Heh! I'm looking forward to that!
btw: I didn't find "popular" on Simpy. Did I overlook something?

greets
Philipp