Re: 3 value logic. Why is SQL so special?

From: Roy Hann <specially_at_processed.almost.meat>
Date: Fri, 22 Sep 2006 15:08:11 +0100
Message-ID: <iMadnbCjtMr1c47YRVnysQ_at_pipex.net>


"JOG" <jog_at_cs.nott.ac.uk> wrote in message news:1158932058.064002.166820_at_b28g2000cwb.googlegroups.com...
> It is surely not a case of whether 3VL is tractable or not, but more a
> question of why the hell would one want to add complexity where it is
> unintuitive and unneeded?

Unintuitive indeed. I just delivered my briefing that I mentioned here a few days ago. Nullability was one of the subjects I intended to address and I did quite a long preamble to it that was good enough that my audience hardly murmured at my claim that one can eliminate the "need" to make many attributes nullable by properly discriminating fact types and using several tables instead of one. But when I went on to argue that many remaining nullable attributes need not be nullable at all it set off a passionate discussion about uncertainty and how to represent tentatively proposed "facts" in the database otherwise.

It soon became clear that most of the (very intelligent) people in the room had a notion that maximum uncertaintly does not lie exactly midway between true and false, but somewhere else entirely! Which seems pretty unintuitive alright.

Roy Received on Fri Sep 22 2006 - 16:08:11 CEST

Original text of this message