Re: 4 the FAQ: Are Commercial DBMS Truly Relational?

From: Kenneth Downs <firstinit.lastname_at_lastnameplusfam.net>
Date: Sat, 09 Oct 2004 10:33:21 -0400
Message-ID: <irs8kc.9nv.ln_at_mercury.downsfam.net>


Laconic2 wrote:

>

>>
>> What are some of the other objections?
>>

> Another objection is this: in mathematics, a relation has no need for
> NULLS. Every tuple is a distinct point in the same space.
> A row with a NULL in it is projected onto a subspace of the original
> space.
>

Hmmm, have never needed NULLS myself, and the customers have never complained. I believe they can always be engineered out. Let the flames begin :)

>
> In my consulting practice, I never tried to distinguish between "a true
> relational database" and Oracle, DB2, or SQL Server. It just wouldn't
> have
> helped communication between me and the client's people. And, to tell the
> truth, I regard Oracle as "close enough for practical purposes". This is
> genuine heresy to the true theoreticians. Ask me if I care.
>
>

Ditto for me, the practical purposes are what pays the bills. Now before Mr. Pascal "debunks" me by pointing out that a rational basis is required to achieve practical purposes, I would ask, if human beings have been using these non-relational systems with an acceptable level of error, and modern civilization has not fallen, is there need for a feedback cycle in which practice affects the selection of mathematical models to use in future engineering?

-- 
Kenneth Downs
Use first initial plus last name at last name plus literal "fam.net" to
email me
Received on Sat Oct 09 2004 - 16:33:21 CEST

Original text of this message