Re: Primary Key Theory Question

From: ben brugman <>
Date: Mon, 29 Mar 2004 16:58:32 +0200
Message-ID: <40683998$0$6575$>

> Correct. Until we get DNA databases, HR data is very often just like
> this.
Identical Twins ?

Often Database designers 'think' that something is unique like the ssn and that 'everybody' has one and just one. In real life, often these rules are not adheared to. And although this should be 'known' at design time it often is not.

This goes for people and ssn. (And not all people have one known birthdate).
For books and isbn. (Not all books have one unique isbn).

So using a surrogate key is often the
best solution in 'real' live.

And even when using surrogate keys, it might happen that you notice one person occuring two times in your database with different surrogate keys.

ben brugman Received on Mon Mar 29 2004 - 16:58:32 CEST

Original text of this message