Re: Distributed foreign keys (was Re: Category Types)

From: Paul Vernon <paul.vernon_at_ukk.ibmm.comm>
Date: Wed, 2 Jul 2003 17:57:48 +0100
Message-ID: <bdv32u$14r6$1@gazette.almaden.ibm.com>

"Lauri Pietarinen" <lauri.pietarinen_at_atbusiness.com> wrote in message news:3EF9EB5A.3060403_at_atbusiness.com...
> So we really end up with
>
> 1) a bunch of datatypes (or domains)
> 2) a set of relation variables (or schemas)
> 3) a set of integrity constraints
>
> Did I leave anything out?
>
> It's pretty simple when you get down to the bottom of it!

Absolutely and therein lies half it's beauty.

Even better than the above however, is that if you consider relation variables themselves simply as constraints on the possiable tuple types in the database variable and that if you have no such constraints then any tuple types are allowed, and you also consider datatypes to be 'external' to the model (and so are not needed to exist in the catalog) then the simplest database possable is (I think) just the empty database.

So all we have in minimum is

D - the database variable
d - the value of the D == {} - the empty set (of tuples)

Regards
Paul Vernon
Business Intelligence, IBM Global Services Received on Wed Jul 02 2003 - 11:57:48 CDT