Re: The Practical Benefits of the Relational Model

From: Paul Vernon <paul.vernon_at_ukk.ibmm.comm>
Date: Mon, 4 Nov 2002 17:43:07 -0000
Message-ID: <aq6bkr$k8u$1@sp15at20.hursley.ibm.com>

"Leandro Guimarães Faria Corsetti Dutra" <lgcdutra_at_terra.com.br> wrote in message news:aq5vc9$6trji$2_at_ID-148886.news.dfncis.de...
> > If so, how do you propose to 'make known to both system and other
> > users' such constraints?
>
> Still waiting for the second part of the article!

You could look in the book.
http://www.amazon.com/exec/obidos/tg/detail/-/0201824590/

> But seriously, the best approximation I know of are the domain, column,
> relation and database constraints.

Agreed. But I would also include the 'constraints' that are the relvar definitions themselves (well, the attribute name & type part of the definition - not the relvar name).

I.e. you cannot insert a proposition into a database if there is not a matching relvar to 'receive it (and,of course it does break any of the 'other' constraints mentioned)

Regards
Paul Vernon
Business Intelligence, IBM Global Services Received on Mon Nov 04 2002 - 11:43:07 CST