Re: Null and Void() - Or, Abandon All Hope Ye Who allow NULLs?

On Wed, Jun 28, 2006 at 02:31:20AM -0700, dananrg(at)yahoo(dot)com wrote:
> Date and Pascal hate nulls. One even goes so far as to say that if
> you permit NULLs in a database, then the results from *every* query
> is suspect. So they turn perform backflips suggesting ways to avoid
> nulls. None, so far, seem appealing.

That is part of the story. Interestingly to me is that the more
stridently they've rejected NULLs, the fuzzier their suggestions have
been about what to do in cases of missing information.

[snip]

> What precisely has Date and Pascal's knickers in such a twist? The
> fact that ad hoc queries from random, unintelligent users could give
> strange results?

What has their knickers in a twist is what twists the knickers of
every other doctrinaire ideologue: that everybody doesn't "just get"
why their way is far superior, drop everything they're doing, and
change over to the True Way™. Their "Eat Grapefruit, You Morons"
<http://www.angryflower.com/getthe.gif> tactics don't do anything to
endear them either.

It's good to read what Date, Darwen & Pascal have to say, but only
once so you can recognize the flavor of bamboozlement when some
impressionable youth (of whatever age) has a "revelation" about How
Databases Should Be®. This way, you can help explain gently that
their "new insight" is neither new nor insightful, and that there are
some good papers <http://www.cs.toronto.edu/~libkin/publ.html> to
read.

Cheers,
D
--
David Fetter <david(at)fetter(dot)org> http://fetter.org/
phone: +1 415 235 3778 AIM: dfetter666
Skype: davidfetter

Remember to vote!