Re: Does 1NF makes sense?
Date: 14 Nov 2002 23:46:12 +0100
Message-ID: <3dd427b4$1_at_news.uia.ac.be>
Juan Pardillos wrote:
>
>I'd like to know what is the sense of defining the first normal form
>(1NF), given that any relation is in 1NF. Perhaps some historical
>reason?.
Historical reasons. Originally when Codd defined the relational model he defined relations as subsets of cartesian products over lists of sets. Since it was not specified what is allowed in these sets a relation could contain anything in its columns, including tuples, relations and other sets. Codd mentioned these possibilities explicitly and then explained why they are undesirable and how they could be removed, resulting in the very first 1NF.
Since then the 1NF has been considered so crucial that many textbooks actually define relations such that they satisfy the 1NF by definition.
- Jan Hidders