NYCPHP Meetup

[Tim Gales]

Mitch Pirtle wrote:
> On 10/10/05, Hans Zaunere <lists at zaunere.com> wrote:
> 
>>So it's a balancing act - as always.  The degree of normalization is
>>generally more of an art than a science, and the "correct" amount is
>>different for every project, and for every database architect.  Get three
>>database architects, and you'll get four recommendations for the correct
>>schema.
> 
> 
> Uh. Hmm.
> 
> *spacemonkey counts with both fingers and toes
> 

It would seem that you should be able to count the
most common degrees of normalization on one hand...
( very few people take things past 5th normal)

Normalization is basically simplifying, if you are
to believe E.F. Codd:

"For this reason (and others to be cited below) the possibility of 
eliminating nonsimple domains appears worth investigating. [see note 4] 
There is, in fact, a very simple elimination procedure, which we shall 
call normalization."

http://www.acm.org/classics/nov95/s1p4.html

There may be a number of ways to simplify a certain schema, and
there may be many opinions about what constitutes a 'good' schema --
but normalization itself is pretty much 'cut and dried'.

Further, if you define the degree of normalization needed as that
degree which will protect the data from (unexpected) anomalies, the
degree of normalization needed is not a matter of opinion either.

A useful guide for the steps involved in normalizing data can be
found at:

http://www.gslis.utexas.edu/~wyllys/DMPAMaterials/normstep.html

-- 
T. Gales & Associates
'Helping People Connect with Technology'

http://www.tgaconnect.com