Re: By The Dawn's Normal Light

From: Ja Lar <jalar_at_nomail.com>
Date: Tue, 26 Oct 2004 09:05:58 +0200
Message-ID: <clksrl$5ps$1_at_news.net.uni-c.dk>


"Alfredo Novoa" <alfredo_at_ncs.es> ......
> "Ja Lar" <jalar_at_nomail.com> ...
>
> >> Candidate keys are at the logical level.
> >>
> >Yes, but "table" is not.
>
> In the context you were talking of course it is.

I'm sorry, but you are wrong. I my context (but perhaps you could not know that): by a table I considered what you call a physical table, ie. an implementation-structure, as opposed to the logical structure of a relation. Of course, if you define a logical table as logical equivalent to a relation, then - by equivalence - a table as well as a relation can be "normal".
In fact, is it not exactly the case that a logical table must be in 1NF to represent a relation?
This resolves the superflous talk about relations at 1NF - it is not relations, but tables, that needs to be in 1NF.

As you see, while elaborating my view, I come to the conclusion that I in fact agree that tables can be talked about on a logical (eg "normalisation") level. Thank you for lifting me! Received on Tue Oct 26 2004 - 09:05:58 CEST

Original text of this message