Re: Is the use of VARCHAR(256) as Primary Keys preferred in Oracle?

From: Paul Brewer <paul_at_paul.brewers.org.uk>
Date: Fri, 22 Nov 2002 20:46:10 -0000
Message-ID: <3ddea001_2@mk-nntp-1.news.uk.worldonline.com>

"Howard J. Rogers" <howardjr2000_at_yahoo.com.au> wrote in message news:cAeD9.81459$g9.229455_at_newsfeeds.bigpond.com...
>
> "Paul Brewer" <paul_at_paul.brewers.org.uk> wrote in message
> news:3ddd6f87$2_3_at_mk-nntp-1.news.uk.worldonline.com...
> > "Howard J. Rogers" <howardjr2000_at_yahoo.com.au> wrote in message
> > news:FYVC9.80802$g9.227327_at_newsfeeds.bigpond.com...
> > > Whatever happened to Boyce-Codd Normal Form, huh??
> > >
> > > Huh?
> > >
> > > Come on... TNF is for wimps.
> > >
> > > Regards
> > > HJR
> > > ;-)
> > >
> > Howard,
> >
> > Have to (yet again) confess ignorance. I am familiar with Uncle Ted's
> > principles, but I have never heard of Mr. Boyce.
> > Is this 4NF and 5NF? (Attribute, Value pairs) and all that stuff? If so,
> I'd
> > theoretically like the idea of designing the entire application in one
> > table, but I might have a slight concern about how it would eventually
> > perform in the real world <g>
> >
> > Regards,
> > Paul
>
> Are you sitting comfortably? Then I shall begin.
>
> A relation is in Boyce-Codd Normal Form (BCNF) if every determinant is a
> candidate key.
>
> Or, as I saw it put more pithily (and accurately, if more inscrutibly):
>
> relation R with FDs F is in BCNF if, for all X ? A in F+
> (where X contains a key for R)
>
> Got that? No??
>
> In other words (hold on to your hats):
>
> R is in BCNF if the only non-trivial FDs over R are key constraints.
>
> From this it follows (follows???!) that there is no dependency in R that
can
> be predicted using FDs alone, and also that if we are shown two tuples
that
> agree upon the X value, we cannot infer the A value in one tuple from the
A
> value in the other.
>
> Dontyajustlurve this stuff??!
>
> I confess to having lifted this lock, stock and several smoking barrells
> from a couple of sites trawled after searching on Google.
>
> All I really know about BCNF in plain English is that it is more
normalised
> than 4NF or 5NF. It is, in fact, the most normalised you can get -it's
> actually theoretically impossible to normalise further than good old BCNF.
> Which is why no-one uses it because attempting to do so would burst
several
> blood vessels in the brain and result in absolutely abysmal performance.
>
> But I like trotting it out at dinner parties, just to impress.
>
> The last dinner party to which I was invited was in 1986. I wonder why?
>
> :-)
>
> Regards
> HJR
>
>

Howard,

Quite straightforward, then.
I'm looking forward to explaining this to the business....

By the way, on the 'sitting comfortably' question, I checked:

pb> lsnrctl status

LSNRCTL for 32-bit Windows: Version 9.2.0.1.0 - Production on 22-NOV-1957 13:45:00

Connecting to (DESCRIPTION=(ADDRESS=(PROTOCOL=IPC)(KEY=EXTPROC0))) STATUS of the LISTENER

Alias                     LISTENER
Version                   TNSLSNR for 32-bit Windows: Version 9.2.0.1.0 -

Service "Mother" has 1 handler(s) for this service.... The command completed successfully

Regards,
Paul ;-) Received on Fri Nov 22 2002 - 14:46:10 CST