If your production database is sufficiently awful, you could export the data into another database server for analysis and queries. I do this with Postgres. My production database(PervasiveSQL) has some crippling weaknesses so I have a python script that slurps the data out and pushes it into Postgres. Then I run some additional sql scripts to renormalize and pre-aggregate some data fields. I then create zope-based web reports that use that second database. If I need to drill-down to specific job details, I can link directly to my production data and the users never see the seam. If you give up and want to pursue something like this, I can send you my scripts.

Although my M$ database is ok, I still consider exporting data to Linux and then querying, so I would appreciate if you send your Zope/Python scripts. Thanks a lot! Rogerio ------------------------------------------------- E-mail enviado pelo Webmail CensaNET http://www.censanet.com.br/webmail