Re: Do Data Models Need to built on a Mathematical Concept?
Date: Wed, 7 May 2003 17:15:16 +0100
Message-ID: <b9be8k$1vs0$2_at_gazette.almaden.ibm.com>
"Neo" <neo55592_at_hotmail.com> wrote in message
news:4b45d3ad.0305061301.5a821adb_at_posting.google.com...
[snip]
> I would say it is even better to see a relation as a a set of things.
> The fact that the thing belongs to the set is all the fact that is
> required.
> The fact that the thing belongs to the set is all the truth that is
> required.
Neo, you did well in explaining your understanding. I understand your
understanding. However your understanding is an incorrect understanding of the
meaning of relations. A tuple in a relation is a statement about things in the
world, it is not (in general) a 'thing' itself. The things are the values in a
domain. I
very strongly suggest you read at least slides 17 to 20 from C. J. Date's
VLDB'02 keynote:
http://www.cs.ust.hk/vldb2002/program-info/keynotes.html#Keynote%202 http://www.cs.ust.hk/vldb2002/VLDB2002-proceedings/slides/S01P02slides.pdf
Regards
Paul Vernon
Business Intelligence, IBM Global Services
Received on Wed May 07 2003 - 18:15:16 CEST