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:cd3b3cf.0110172223.26b3b5e7_at_posting.google.com...
> "Daniel Poon" <spam_at_spam.com> wrote in message
news:<1003222318.886981_at_kang.qonos>...
> > Answering my own question here, I seem to remember that when I first
studied
> > logic, we had propositions such as:-
> >
> > Greek(Socrates)
> > Bipedal(Socrates)
> > Rational(Aristotle)
> > Man(x)=>Rational(x)
> >
> > So, yes, it does make sense to have predicate logic without set theory.
>
> You have not answered your original question. The relational algebra
> is set theory. The relational calculus is predicate calculus. While it
> might make sense to have relational calculus without set theory, it
> makes no sense to have relational algebra without set theory.

Im not sure what you mean by relational algebra vs relational calculus. Predicate calculus is not set theory.
rdbms (the one I know of) are a simplification of predicate calculus.

> > Do you really understand set theory and predicate logic? I have
confessed
> > that I am no expert in the field, even though I studdied
> > Maths/logic/foundations of maths for 3 years. Because of this discussion
> > though, a lot of it is comming back to me ;-)
>
> I understand them well enough. Do you understand the relational model?

Well I studdied predicate calculus and set theory. And Ive used rdbms. I haven't formally studdied the relational model if thats what you mean.

> > The relational model DOES NOT contain set theory. Set theory is used
> > underneath to explain how it all works. But you can't manipulate sets
> > directly.

> The relational model builds on set theory. The relational algebra is
> direct application of set theory just as the relational calculus is
> direct application of predicate calculus.

The whole of maths today is built on the foundation set theory, so saying that rdbms is build on set theory doesn't say very much, since mathematitians would argue that EVERYTHING is built on set theory.

> I do not know what it would mean to "contain" set theory. Even set
> theory has trouble with this.

I think in our finite world of desktop computers, we don't have to worry too much about these philosophic problems ;-)

Daniel Received on Thu Oct 18 2001 - 11:59:08 CEST