Re: What does this NULL mean?

"dawn" <dawnwolthuis_at_gmail.com> wrote in message news:1134694213.377635.258420_at_g43g2000cwa.googlegroups.com...

> only slightly? Embrace your heresy. One thing that I think is very
> important is a change in how the RM is taught at the undergrad level.
> It is taught as if it was handed to us as truth by God herself. There
> are many ways to model propositions for data processing, the RM being
> only one of them (and not the most flexible IMO).

I think this is close to the heart of the matter. Not this particular topic, but the underlying matter that we always seem to come back to. I would ask if the axe is sharp enough yet, but I don't wan't you to misread that phrase like last time.

First, I find it extremely odd that RM would be taught as revealed truth. When I taught database design for DEC, one of the points that was made explicit in the opening lesson was that, when it comes to design, there is no one single right answer. Rather, there are many answers that deliver varying degrees of satsifaction, and different dimensions along which to measure satisfaction. Of course, I'm talking about "how to design a database" not how to choose a data model. And the question of whether the RM was apropos or not was already made by the choice of the student to attend this particular course.

We had other courses in how to design databases for VAX DBMS (a CODASYL network model DBMS). I taught Rdb, which was based on the RM.

But that's still not the heart of the matter. The heart of the matter is this: "model propositions for data processing". and the key word here is "processing".

I think that Codd and Date wrote about the use of the RM (or, more precisely, the RDM) in the context of data storing and data sharing, and not in the context of data processing. It's not clear to me that you want to model data the same way for processing that you want to model data for storing and sharing.

If you're storing, but not sharing, that's another story. But is that story about "databases" or not? Received on Fri Dec 16 2005 - 10:05:52 CET