Bess Sadler wrote:

> Hi, Andrew. Since this will be an all-day event, the session would be
> starting first thing in the morning on Feb 27. I'm thinking 9am, but
> I haven't confirmed that with anyone else. I'm just flying by the
> seat of my pants here.

I wouldn't be able to make this then due to time constraints.

> That way you can use solr / lucene for search, faceted
> browse, etc, and your XML database only for known item retrieval,
> which it is generally able to do without performance issues.

I am doing something similar except I am using my file system as my
database for pulling the full marcxml records.  This offers little
overhead as possible.  Now think about the possibilities of using
something like lucene or postgres as your filesystem.  There are many
groups working on these such filesystems for years.

> I'm
> hopping up and down waiting for someone to take this approach with an
> ILS, so please come and show us what you've got!

I have proposed a talk on my trials and tribulations of developing this
at this years code4lib conference.  If it is accepted I will share all
the gory details.

BTW, have you played with Hadoop?  I guess it's something like the
open-source attempt to google's search algorithm.  I would be curious
about implementing hadoop across a few servers to store the marcxml records.

Andrew