Re: Dawn doesn't like 1NF

From: Alfredo Novoa <alfredo_at_ncs.es>
Date: Thu, 07 Oct 2004 16:45:35 GMT
Message-ID: <416571ce.25511796_at_news.wanadoo.es>


On Thu, 7 Oct 2004 11:42:21 -0400, "Laconic2" <laconic2_at_comcast.net> wrote:

>When are you going to stop with this, Alfredo??? It's getting old. Of
>course there are references to 1NF in the 1970 paper! The term used in the
>paper is "normal form".

No, 1NF is not that. What Codd wrote about normalization in the 1970 paper is now obsolete. The paper has historical interest only.

A modern formulation of the 1NF could be something like this:

All relations are in 1NF.

>> The problem is that "simple" and "atomic" are not precise terms.
>
>I may be imprecise in my wording. Codd tried to be precise in his wording.
>I believe he succeeded.

You are wrong. He failed and it is explained on the books.

You should read Date's writings about 1NF.

>> For instance: Are strings simple or complex?
>
>It depends. From the point of view of the DBMS, they are simple. From the
>point of view of the SUBSTRING function, they are complex. That doesn't
>mean that the definitions are vague.

Your point is very weak and fuzzy.

We could say the same about a relation value: it is simple from the point of view the DBMS (whatever that means!) but complex from the point of view of the tuple extractor function.

You are trying to prove that the terms are not vague using a lot more vague terms.

Regards Received on Thu Oct 07 2004 - 18:45:35 CEST

Original text of this message