Re: Codd provided appropriate mathematics ... (was Re: Relational and MV (response to "foundations of relational theory"))

"Marshall Spight" <mspight_at_dnai.com> wrote in message news:lcT%b.409798$I06.4477771_at_attbi_s01...
> "Dawn M. Wolthuis" <dwolt_at_tincat-group.com> wrote in message
news:c1ocbm$ldq$1_at_news.netins.net...
> > Your last statement, Eric, is really the point. I think the original
> > idea was that relational database theory was tied to mathematical
> > relations. It is now at the point where statements that are true of
> > mathematical relations such as the fact that relations are ordered,
> > are now considered "obviously" untrue in the database relational
> > model. Ah well, such is the nature of language. --dawn
>
> If I may make a correction: in mathematical relations, the attributes
> are ordered instead of named. The elements are unordered in both
> models.
>
> I consider this difference relatively minor.

Yes, the elements of the ordered set (relation) become attributes in an unordered set of elements that include names. There are other relational database theories that actually keep the relation a relation. The fact that the theory that likes exclusive use of the term "relational database theory" doesn't actually employ relations is not a major point on the one hand, but confuses the language in a significant way. It is really a pain in the neck when I refer to a relation used for data modeling and am told it is not a relation -- UGH!!

So, is there agreement that THE RELATIONAL DATABASE MODELING THEORY or whatever you want to call it DOES NOT MODEL DATA AS RELATIONS? I do. PICK does. What in the world am I supposed to call the relations that ARE relations but are not relational database theory relations (which are NOT relations)? Can you see why this is such a pain when trying to discuss various models? --dawn Received on Sat Feb 28 2004 - 18:34:10 CET