Re: Codd provided appropriate mathematics ... (was Re: Relational and MV (response to "foundations of relational theory"))
Date: Fri, 27 Feb 2004 12:32:05 -0600
Message-ID: <c1o2fn$ekj$1_at_news.netins.net>
Amen, Tom -- I mean, yes, you are right on both points. From a mathematical perspective there is no reason that a relation representing a proposition (or portion of one) needs to be in 1NF. However, when one speaks of a "relational database" many people are no longer talking about the mathematical relations nor to all possible relational models, but about a particular religion related to these relations.
I do appreciate zealousness and passion, and religion certainly has a place in discourse, but I have tried to make it clear when I use intuition to derive an opinion, compared to other more acceptable forms of logic. It is important to see with any model where the model is a good fit and where it is not -- where we use logic to derive information and where we have leaps of faith. It is unfortunate that our institutions of Higher Ed often teach relational databae theory as if it were "the Truth". Cheers! --dawn
"Tom Hester" <$$tom_at_metadata.com> wrote in message
news:bac94$403f841e$45033832$27879_at_msgid.meganewsservers.com...
>This is just plain false. There have been many relational models proposed
that are not 1NF. See for >example, ABITEBOUL, S.; BIDOIT, N. Non first
normal form relations to represent hierarchically >organized data. In: PODS,
1984. Proceedings. . .Note the date: 1984. This is supposed to be a
database >theory board but it seems more religious than theoretical!
Received on Fri Feb 27 2004 - 19:32:05 CET