On Wed, 20 Jun 2007, brian wrote:
The former uses a primary key across both columns to enforce a unique
constraint. In the latter, you have a seperate ID column, which does not
enforce that constraint. And you have to ask yourself if you'll ever be
referencing that ID column for anything at all. I doubt i ever would.
Generally, you'd be using this to relate rows from a more generalised
table using either the club ID or the user ID. I can't see how having a
seperate serial ID column would be useful for any kind of select.
Also, the reason for a third, M-M, table is to relate multiple players and
multiple clubs. If you think of the logic involved, your third table has
only one row for each player-club combination. Therefore, each row is unique
by definition and a surrogate key adds no value.
Rich
--
Richard B. Shepard, Ph.D. | The Environmental Permitting
Applied Ecosystem Services, Inc. | Accelerator(TM)
<http://www.appl-ecosys.com> Voice: 503-667-4517 Fax: 503-667-8863
