>-----Original Message-----
>From: Douglas McNaught [mailto:doug@mcnaught.org]
>Sent: zondag 28 januari 2007 16:29
>To: Joris Dobbelsteen
>Cc: John Meyer; pgsql-general@postgresql.org
>Subject: Re: [GENERAL] counting query
>
>"Joris Dobbelsteen" <Joris@familiedobbelsteen.nl> writes:
>
>> What would have been better without surrogate keys all-over:
>> * Easier to write complex queries with much fewer tables to
>be queried.
>> * Much faster query performance, as fewer tables need to be
>referenced.
>> * Better integrity enforcement with simple foreign key constraints.
>
>Not this debeta again. ;)
>
>Surrugate vs natural keys shouldn't make a difference in how
>many yables you have--they depends on the degree of
>normalization. Sounds like you denormalized your database and
>happened to eliminate surrogate keys at the same time. Using
>that to say "surrogate keys are bad" is kind of misleading.
I have perhaps formulated it quite extreme. It was not intended to take
such a extreme stance. I appologize if it will lead to such a
discussion. I'll try to do it a bit more careful next time.
My point is only, be careful with surrogate keys and try not the use
them for everything. In my sole opinion I see them getting used too
much. But then again, what is good and wrong will always be subjective.
So perhaps <quote>What would have been better without surrogate keys
all-over<quote> should have been "My database where I extremely overdid
it with surrogate keys".
Lets leave it to this.
- Joris