Re: A question for Mr. Celko
Date: 17 Jul 2004 23:38:44 -0700
Message-ID: <72f08f6c.0407172238.41740e15_at_posting.google.com>
> EXACTLY! This means that I can use characteristic functions on
> columns, and therefore on rows, and therefore on table. No recursion,
> no self-references, no indeterminable sets, etc. Get a slightly
> advanced book on sets and look all the weird sets that are defined so
> that you are not sure what the elements are.
The only definition I can find for characteristic function is a fancy word for the in operator. Is this the meaning you are intending? If so, how could you invoke it on an attribute without relation-valued attributes?
As for no recursion, self-references, and indeterminable sets, it sounds to me like this is just unnecessarily restricting the power of the language, a technique extremely common in today's systems, and one of the reasons relational languages get such a bad rap. Can you give a specific case where allowing relation-valued attributes would result in a wrong answer? Are you suggesting that allowing list, tuple, and relation-valued attributes somehow makes the language inconsistent (inconsistent in the formal sense meaning it would be possible to derive contradictory theorems)? Received on Sun Jul 18 2004 - 08:38:44 CEST