Ian Bicking: the old part of his blog

Re: Schevo and Durus

For what it's worth whenever I've dabbled with OODBs I haven't been able to quickly answer questions which involve aggregation, and I always want to do that to a greater or lesser extent. You know, the sort of queries like "Give me all of the transactions in the third quarter last year over five thousand dollars from suppliers in Western Australia"

This is the strength that catapulted relational databases over and above their network and hierarchial competitors in the 70's and 80's. Removing the need to re-organise your data when you want to look at it in a slightly different way made a big difference in the real world where change is the only constant. When object databases crack it (perhaps with a consistent query language) then I confidently predict adoption will skyrocket.

There is a lively thread on this very subject on the db-sig mailing list, look for 'Pycon2005 and database divide' at this page:

http://mail.python.org/pipermail/db-sig/2005-March/thread.html

Comment on Schevo and Durus
by Andy Todd