Re: terminology

U-gene wrote:
> Just a remark
>
> Marshall wrote:
>
> > Most basically, a relation is a subset of a product of sets.
> > In the c.d.t. context, we mean a little more than that, because
> > of keys and possibly constraints, and because each of the sets
> > has an associated name, but that's the basic idea. A particular
> > relation is a value.
>
> I've got a old book of David Maier "The theory of Relational Dataase"
> (IMHO it's only really theory book on subj). There is very interesting
> definition (which can be called "orthogonal" to usual one) of what
> relation is in this book. I'll try to translate it back form Russian to
> English :)
>
> "...Relation r, which has schema R{A1, A2 ... An}, is a finitesimal set
> of the mappings {t1, t2 ... tn} from R to _D_, where _D_ is union of
> domains D1, D2.... Dn , and each mapping t(Ai) belongs to Di. ..."
>
> I don't know if I could translate it correctly. It is interesting
> becouse it define indissoluble relationship between schema and body of
> relation. There is NO something special in schema or in body - they are
> just set(s) of values becouse each Ai is name(i.e. value) and each
> t(Ai) is (scalar) value. According to this(and all the more ) -
> relation is value.
In essence, what you are just trying to establish for RDM what has already established in math for centuries: an *indissoluble* (to use your term) relationship between variable and value. No need to spend hours or ellaborate on that. I prefer the mathematical inspired concept of value holder for variable. In RDM, an attribute is nothing but an elementary relvar, result of an intersection between the relvar defined domain and N single value domains. Received on Thu Jun 22 2006 - 08:18:42 CEST