Re: Database design

From: x <x_at_not-exists.org>
Date: Thu, 23 Feb 2006 17:12:57 +0200
Message-ID: <dtkjdq$hc4$1_at_emma.aioe.org>


"Alexandr Savinov" <spam_at_conceptoriented.com> wrote in message news:43fdc083$1_at_news.fhg.de...
> x schrieb:
> > You say a relation is flat because the attributes of a single tuple are
all
> > at the same "level" like (a1,a2,...,an) in R(A1,A2,A3,...,An) instead of
> > something like (((a1,a2), (a3,a4)), a5, (a6,a7,a8),....) so the tuples
are
> > flat, not the relations.

> Not precisely, but your interpretation is very close to what I tried to
> say and I think quite compatible (concerning tuples).

> But actually I do not insist on my interpretation of the term "flat".
> I simply describe what I find natural.

It would be useful if this "flat" term is not just decorative. The idea is that the hierarchy is in the eyes of the beholder. Anyway the RM evolved over time.

>The idea is very simple. If we have a set with
> elements then I tend to consider it flat because I do not see anything
> behind these elements and over this set.

>A single relation is precisely
> such a set without more specific and more general levels (if you
> consider only one relation - not many).

A relation is not a set.
If it were then the only operations available would be the set operations. Received on Thu Feb 23 2006 - 16:12:57 CET

Original text of this message