Re: Do Data Models Need to built on a Mathematical Concept?

From: Marshall Spight <mspight_at_dnai.com>
Date: Sat, 26 Apr 2003 02:14:57 GMT
Message-ID: <BYlqa.85410$Si4.67592@rwcrnsc51.ops.asp.att.net>

"Ian" <kellizer(nospam)@hotmail.com> wrote in message news:av8qa.611$oA1.85469_at_newsfep2-gui.server.ntli.net...
> I think that's a widely inaccurate notion, that if data models don't follow
> a mathematical concept, they are not "correct". Data model follow rules, If
> the model conforms to the rules, it is correct.

Well, what I'd call a data model is a set of rules. It sounds like perhaps you're calling the implementation of the rules in code the "data model."

It seems to me that if the data model, the rules, or whatever, are built on sound mathematical foundations, you've got something sound. If it was just thrown together in order to be easy to implement, (the usual case) then what you've got is anyone's guess.

> You talk about the
> implementation of data models as a deciding factor is the design?

I didn't quite understand this sentence, but I'll note that I'm just talking about model, not implementation; they are separate.

See also Costin's post, which says a lot of interesting and important things really well. (That is, better than I could have.)

Marshall Received on Fri Apr 25 2003 - 21:14:57 CDT