Re: S.O.D.A. database Query API - call for comments

Simon Brooke wrote:
>
> on Saturday 28 April 2001 21:53, Lee Fesperman wrote:
>
> > Object-oriented is a bunch of half-baked concepts with no theorectical
> > foundation. There is no way to prove these concepts except
> > empirically. Relational DBMSs, on the other hand, are built on
> > fundamentally sound principles.
>
> No, sorry, that's really, really not true. Just as the relational model
> has a sound basis in the mathematic of relational algebra, so object
> oriented design has a sound mathematical basis in category theory.

I 'really, really' have doubts about that. Relational is mature, proven technology. I've never even heard of this in OO, much less seen it in use.

> This is not to knock the relational model; it is my opinion, as it
> seems to be yours, that at the current state of the art the relational
> database is the most effective technology for storing many types of
> data in common use. And it is my opinion (whether or not it is yours)
> that no computing formalism without a sound mathematical basis can ever
> really be successful. But the strength of the relational database's
> position should be such that there's no need to knock other
> technologies.

If we were discussing different programming languages, I might agree with you. However, much of the discussion was about database models. It's perfectly reasonable to knock a technique for programming being arbitrarily applied to database using a hodge-podge of long-ago discredited techniques.

As mentioned earlier, COBOL programmers want database access to be more like COBOL, thus was born CODASYL.

-- 
Lee Fesperman, FFE Software, Inc. (http://www.firstsql.com)
===================================================================
* Check out Database Debunkings (http://www.firstsql.com/dbdebunk/)
* "The Forum Where Database Matters Are Set Straight"