Re: A Normalization Question

> > ... the limited self-constructed matrix is RM.
>
> Maybe, but *your* model is *your* limited self-constructed matrix.

All data models (RM, TM,...) are constructed "matrix" each with a limited scope to represent things. I contend that RM's scope is a subset of TM's. But there are alway pros and cons. Within RM's smaller scope, things will be easier and more efficient to represent and manipulate. The further beyond RM's scope, things will become less and less practical, if not impossible.

For example, while RM can represent strings and symbols without redundancy at the logical level, it is rather impractical, but not impossible.

The following example (carried out far enough), is even more impractical, if not nearly impossible. Try representing the following in RM without NULLs or redundant things:

thingA: john take final
thingB: john receive 95
thingC: thingA cause thingB 

In effect thingC would be:
(john take final) cause (john receive 95) Received on Tue Jul 20 2004 - 23:56:01 CEST