User-defined database types

Invitation for reader comment

How well do UDTs (user-defined types) work in production databases?

While I've worked in data management for decades already, I've never had the experience of running a production database that makes what I consider full use of UDTs, let alone fully design and implement one myself. Type theory is one of the most secure rocks on which to ground computing practice, and it ought to provide considerable leverage to data architects who care about the correctness of their repositories. On the other hand, although SQL Server, Oracle, DB2, and other common RDBMSs support UDTs, SQL itself doesn't (yet) specify a standard for UDTs, so many organizations avoid them. In fact, they're arguably less prominent in current documentation than they were in 2000 or so. I've never made the weeks it would take to experiment with them on the scale that interests me. What do you know about UDTs in database practice? Have you used them at all? Is there a particular forum or mailing list that treats them in an adult way at, say, the level of Gio Bejarasco's Vector project?

This story, "User-defined database types" was originally published by ITworld.

Related:

Copyright © 2009 IDG Communications, Inc.

  
Shop Tech Products at Amazon