Re: The Fact of relational algebra (was Re: Clean Object Class Design -- What is it?)

"Bob Badour" <bbadour_at_golden.net> wrote in message news:8iZx7.2887$uS1.94257252_at_radon.golden.net...
> "Daniel Poon" <spam_at_spam.com> wrote in message
> > Look at this bit of logic "Socrates is a man"
> >
> > i.e. Man(Socrates)
> >
> > It is not asserting anything about a string of ascii characters or any
> other
> > kind of key.
>
> It is not asserting anything about a location or an OID either; however,
the
> unqualified string (or symbol) "Socrates" suffices to identify the
> individual to the participants of this discussion. Do you have a point?

Oh please. I haven't even mentioned OIDs. Why have you brought them into the discussion?

> > Socrates is also a member of the Universal set, and a lot of other sets.
> The
> > fact that he belong to these sets is independant of his existance.
>
> Agreed. The relational model allows this. Again, do you have a point?

And how would you implement the Universal set in an rdbms ;-)

The point is not that you can express one in another. The point is that the fundamental data representations are different.

Heck, in my oodbms I could create a class called Tuple, a class called row, and a class called SQLquerry, and build an rdbms in my oodbms. But thats not the point either.

Daniel Received on Tue Oct 16 2001 - 10:41:19 CEST