Re: O'Reilly interview with Date
Date: 3 Aug 2005 21:06:53 -0700
Message-ID: <1123128413.356263.171690_at_g43g2000cwa.googlegroups.com>
dawn wrote:
>
> I like theory based on practice, but that is not the same thing as
> proving that other theories are not as good.
Goodness isn't something you prove.
Proofs can demonstrate that certain mathematical qualities will always hold within a closed system. You can, for example, prove soundness, or self-consistency, or completeness against a given set of criteria. But you can't prove goodness because you can't measure it.
> I haven't seen such a
> proof from any relational theory camp, just bad-mouthing of tree or
> di-graph models, dismissing them as "hierarchical" and "network". It
> seems like buzz and not logic.
Any such evaluation is going to be based on value judgements. Consider static typing vs. dynamic typing. Lots of proponents on both sides. Good arguments for each. But the central difference boils down to the fact that the static typing camp prefers to detect errors early and is willing to pay some price in flexibility, and the dynamic typing camp wants to maximize flexibility and is willing to pay some price in deferring finding of errors. You can't "prove" which choice is correct; you can merely have a preference.
Some folks like relational, because they like the algebraic properties, the declarativeness, etc. They value these things highly. Maybe you don't, and you different value function produces some other result. But there's no "logic" to be had in the debate. One has one's own set of weights to apply to various features, and one "does the math." The weights are axiomatic. The result isn't a proof, it's a solution to a set of requirements.
Marshall Received on Thu Aug 04 2005 - 06:06:53 CEST