Re: System (Hidden) Keys

From: Dave Boswell <bd860_at_FreeNet.Carleton.CA>
Date: 1995/06/01
Message-ID: <D9H1nH.ID4_at_freenet.carleton.ca>#1/1


Randy Dewoolfson (randyd_at_cais2.cais.com) writes:
> We are about to have a massive debate in the company I work with.
> The discussion will be about primary keys that are system generated,
> (like by a sequence) vs. a set of 'regular' fields that can be used
> as unique primary keys.
>
> What I'm looking for is basically a list of PROs and CONs about
> defining Primary keys in these two ways.
>

[ stuff deleted ]

There have been various articles published in DATABASE Programming & Design and DBMS magazines, both published by Miller Freeman Publications, in the last year or so. One that you might find useful, depending on which side of the fence you are on, appeared in May 1995 DATABASE According to Date column. Its titled "Say No to Composite Columns". Although its not directly responding to the key debate it does describe some of the problems of composite columns/keys. In this field it's sometimes useful to have the writings of Mr. Date and his long history in the database field on your side. By reading his work, and that of the people that disagree with him, you should be able to develop this list in very short order.

A direction, not an answer. Dave Boswell. Received on Thu Jun 01 1995 - 00:00:00 CEST

Original text of this message