Re: The Practical Benefits of the Relational Model

From: Paul Vernon <paul.vernon_at_ukk.ibmm.comm>
Date: Fri, 4 Oct 2002 18:24:45 +0100
Message-ID: <ankj7e$13a6$>

>>> God forbid nesting/unnesting!?
>>Fully allowed.

>It is? So how is the 1NF defined then in the relational model? Or is
>now a new "*the* relational model". :-)


Scientists don't currently seem to work that way. I.e. AFAIK there is not a single body that administers *the* relational model, unlike say for SQL.

Nor am I aware of any body centrally defining the Theory of Evolution, or even Gravity.

Now whether such bodies would be a good idea would make an interesting debating point, but not one particularly on topic for this group I suggest. Suffice to say that I'd like to see it occur, but only if different bodies were forced to compete on getting the best formulation of the same theories.

Back on topic. the Chris Date Relational Model (TM?) has allowed relation valued attributes since at least 1994. The version of Darwen & Date's Manifesto in his DB writings book 91-94 explicitly allows relation valued attributes, but explicitly does 'not exposure NF-squared'.

Now you might disagree with that particular squaring of a circle, but I must admit I agree with the two of them.

Why should Relation values be explicitly the SINGLE datatype out of the universe of all possible database that are disallowed from being attribute values?

And besides that, they are darn useful in cases like the System Catalog.

Paul Vernon
Business Intelligence, IBM Global Services Received on Fri Oct 04 2002 - 19:24:45 CEST

Original text of this message