Re: Do Data Models Need to built on a Mathematical Concept?

From: Costin Cozianu <c_cozianu_at_hotmail.com>
Date: Fri, 02 May 2003 15:09:37 -0700
Message-ID: <b8uq55$dut5f$1_at_ID-152540.news.dfncis.de>


Neo wrote:

>>>Just as Newton's Theories were
>>>displaced by Quantum Theories, so will the Relational Data Model.
>>>
>>
>>It was not exactly displaced, was it?  It was just refined...

>
>
> Yes and No. In some applications, Newton's methods provide results
> that are "close-enough" and a lot easier to calculate. In other
> applications, Newton's methods produces results that are "way-off" and
> useless.
>
> Similarly, in some applications (which at the present seems to be the
> majority), the relational methods provides results that are
> "close-enough" and infact better than most other methods.
>
> But, there are applications (which at the present seem to be a
> minority) where the relational methods results are not satisfactory
> such as highly variable/complex data structures that could be thought
> of as graphs. IMO, the model used by the human brain is some
> refinement of the relational model.

I guess you are first of all very confused as to what the relational model is, therefore you probably are extremely unqualified to assert with no arguments: that "the relational methods results are not satisfactory such as highly variable/complex data structures that could be thought of as graphs" .

You need to learn is that a graph is equivalent to a binary relation, and incidentally the relational model is very well suited for guess what: relations.

As to "the model used by the human brain", if you knew that, you'd be a Nobel prize winner, and you wouldn't waste your time trolling comp.database.theory :) Received on Sat May 03 2003 - 00:09:37 CEST

Original text of this message