Re: By The Dawn's Normal Light

From: Dawn M. Wolthuis <dwolt_at_tincat-group.comREMOVE>
Date: Wed, 27 Oct 2004 19:48:48 -0500
Message-ID: <clpfm2$ou$1_at_news.netins.net>


"Laconic2" <laconic2_at_comcast.net> wrote in message news:pYWdnQw_rL7Mix3cRVn-hg_at_comcast.com...
>
> "Dawn M. Wolthuis" <dwolt_at_tincat-group.comREMOVE> wrote in message
> news:clovf6$plp$1_at_news.netins.net...
>
> > It seems that every first course in databases still teaches how to put
> data
> > into the old def of 1NF, and the pervasiveness of SQL-92 perpetuates
that
> > too. So I'd prefer to reitre the term 1NF so we stop teaching it with
the
> > old def. It adds nothing to the discussion. Then we can rename the
upper
> > level NF's so they are not numbered, but named. --dawn
>
> First off, it's not clear to me that the body of IT "Referees" or
"Experts"
> or whatever they are have indeed reached a consensus that Date's
alteration
> is to be commonly accepted. It is certain that there is no consensus on
> that view in this forum. So, if "we are IT", then we are still in the
> process of reaching consensus.
>
> Second, I find it ironic in the extreme that the great theorists of data
> and databases should have mismanaged the namespace so badly in the case of
> naming normal forms. First, the numbering of the normal forms as "1",
"2",
> "3" with the definitions as successive additions to the previous ones,
> basically suffered from the same defect as using a surrogate key, only
> worse. This surrogate key was overloaded because in addition to carrying
> the order of discovery, it also carried the sequence of progressive
> definition.
>
> So, when 4th normal form was discovered and named, before Boyce-Codd
normal
> form, they were already up the creek. The only alternative would have
been
> to call Boyce-Codd normal form something like "3.5 normal form". And we
all
> know what would have happened then!
>
> Next, along comes Date, and redefines "relation" in such a way that
"all
> relations are in 1NF". (Is this really what happened?) Trouble is, the
old
> definition of relation is still current in the world of mathematics. So
now
> we have the "homonym problem" in the namespace as well, if we include
> mathematics and IT in the same universe of discourse. And the RDM was
> supposed to unify mathematical and "informatical" disciplines! Great
> example for the unwashed masses, guys!

Yes, good points. Let's pretend we are fixing it and come up with an alternative. Then we will need to sell it. The column writers who listen in on this discussion could help us with that. cheers! --dawn Received on Thu Oct 28 2004 - 02:48:48 CEST

Original text of this message