Re: 1GB Tables as Classes, or Tables as Types, and all that refuted

From: Ja Lar <ingen_at_mail.her>
Date: Wed, 24 Nov 2004 19:52:30 +0100
Message-ID: <41a4d881$0$205$edfadb0f_at_dread11.news.tele.dk>


"Alfredo Novoa" <alfredo_at_ncs.es> skrev i en meddelelse news:41a4bcf4.21479484_at_news.wanadoo.es...
> On Wed, 24 Nov 2004 16:26:18 +0100, "Ja Lar" <ingen_at_mail.her> wrote:
>
>>> An attribute of a relation can not have another relation as its
>>> domain. This is an absurd.
>>
>>I recommend you Date's Intro ...(8th ed), Part II/The Relational Model p.
>>152: Relation-Valued Attributes.
>>You might want to inform him about the absurdity :-)
>
> No, you are missing the point. Relation valued attributes don't have a
> relation value as their type (domain), they have a relation type as
> their type.

OK, I acknowledge that.
So the set of values specified by the type is (in this case) a set of relations.

But the set of values that a relation is ("contains") cannot be mapped to the set of values that a type is ("contains"), ie. "a relation cannot be a type", if I understand you right ? Received on Wed Nov 24 2004 - 19:52:30 CET

Original text of this message