Re: WWW/Internet 2009: 2nd CFP until 21 September x

From: Walter Mitty <wamitty_at_verizon.net>
Date: Sun, 16 Aug 2009 03:13:43 GMT
Message-ID: <HPKhm.2338$Jg.2098_at_nwrddc01.gnilink.net>


"paul c" <toledobythesea_at_oohay.ac> wrote in message news:FfHhm.41439$PH1.12688_at_edtnps82...
> Bob Badour wrote:
> ...
>> Each of those unnamed, implicit, ephemeral tables ...
>
> Ephemeral, that's for sure, the anti-matter of the db world. Even before
> I learned any of the theory, I had a feeling people who wanted outer join
> couldn't keep their eye on the ball if they tried and believed they could
> ride their bicycles backwards even if they'd never actually tried to.
> Don't know what Codd might have said about it but whatever it is, if it is
> anything but a brief mental mishap that is forgotten when one sobers up, I
> can't imagine a join that isn't a concrete inference, too mystical for me.

When you have ot produce reports, outer joins can be useful. Trust me.

The fact that an outer join on two relations does not yield a relation should be a clue to those who want to work out the operation of a database engine in terms of mathematics. To those of us who are simply using tools like SQL to build systems, the mathematics sometimes helps and sometimes doesn't.

Having said that, I agree with those who say that the reason relational databases have endured so long and worked so well is that they are more nearly grounded in sound mathematics than most of the stuff in computer science. Received on Sat Aug 15 2009 - 22:13:43 CDT

Original text of this message