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

On 30 Nov 2004 15:54:37 -0800, rafe_at_cs.mu.oz.au (Ralph Becket) wrote:

>A type is a set of values supporting the same operations.
>Sometimes `type' is further refined to require that all
>values in the type have the same representation.

Such refinement is clearly a mistake because it mixes the model and implementation levels.

A value is a member of a type independently of the way we are representing it.

Then arabic number 20 represents the same value as the roman number XX and such value is an integer value among other things.

>I'm guessing that Rene is using simple and complex to mean
>primitive and compound respectively.

I agree.

Regards Received on Wed Dec 01 2004 - 17:12:32 CET