Re: Logical equivalence of simple and complex types under the relational model?

From: Alfredo Novoa <anovoa_at_ncs.es>
Date: Thu, 02 Dec 2004 02:04:14 +0100
Message-ID: <g7qsq0ttkanetjaj9sh2oi9bsfblf1lq9f@4ax.com>

On Wed, 1 Dec 2004 20:58:02 +0100, "Rene de Visser" <Rene_de_Visser_at_hotmail.com> wrote:

>Does this mean that the later debates about whether complex values should be
>allowed in databases,
>were debates (and papers) about nothing?

There are debates about nothing. It does not make sense to distinguish between simple and complex scalar types.

>Or is there some definition under which these debates make sense?

Databases should have rich, powerful and well designed scalar type systems and that's all.

>Though I find it hard to imagine any definition under which the above
>'transformations' can not be applied.

That's the problem. We can not find a sound definition.

Regards Received on Wed Dec 01 2004 - 19:04:14 CST