Re: SQL Server and User Defined Functions

From: Paul G. Brown <paul_geoffrey_brown_at_yahoo.com>
Date: 19 Mar 2002 17:05:39 -0800
Message-ID: <57da7b56.0203191705.394574ce_at_posting.google.com>


"runstein" <runstein_at_biotronik.com> wrote in message news:<u9eultq22qgde0_at_corp.supernews.com>...

> Besides the obvious portability issues, could you guys give me your thoughts
> on udf's? Am I being paranoid or are there some real reasons for my
> hesitancy?

  Most DBMS products have UDFs now, although their flavor varies. They're quite compelling in situations where the presence of the UDF in the query language reduces or eliminates the data that needs to be passed between the SQL and a procedural language (even a stored procedure). They make a lot of data modeling sense in that they let you do a better job of achieving a modular database schema design.

 In general I would avoid writing UDFs that contain SQL: instead, use the SQL as a mechanism for working with large numbers of UDFs.

  They're really kind of cool.

  And if you're worried about portability, what the heck are you using SQL Server for?? ;-) Received on Wed Mar 20 2002 - 02:05:39 CET

Original text of this message