Re: relations aren't types?

From: Mikito Harakiri <mikharakiri_at_iahu.com>
Date: Tue, 30 Dec 2003 09:46:12 -0800
Message-ID: <FZiIb.6$7S6.111_at_news.oracle.com>


"Bob Badour" <bbadour_at_golden.net> wrote in message news:XeCdnbxDlZnbh2yiRVn-vA_at_golden.net...
> "John Jacob" <jingleheimerschmitt_at_hotmail.com> wrote in message
> news:72f08f6c.0312291844.48dd11fc_at_posting.google.com...
> > > The relation value in the attribute for each tuple is
> > > a single value with defined operations. How does that differ from an
> > > integer?
> >
> > Types are not atomic, values are.
>
> How does that answer the question about the difference between a relation
> value and an integer value?

I beg to disagree. This was the only sentence in the whole discussion that made sence to me.

Consider the "Group" type. Now, some groups are decomposable into direct products, like dihedral group D_3 (aka symmetric group S_3), while the others, like cyclic group Z_3, don't. I'm tempted to call Z_3 atomic, and S_3 composite value, then.

Likewise, some integers can be represented as products of smaller numbers, while the others - prime numbers - don't. In some cases, values are decomposable for the whole domain, e.g. complex numbers. Every complex number is a direct product of real and imaginary component.

I'm not sure if this elementary mathematical perspective onto atomicity makes any sence from programming perspective, but it is relationists who emphasise mathematical soundness, right? Received on Tue Dec 30 2003 - 18:46:12 CET

Original text of this message