Re: A Normalization Question

From: Jan Hidders <jan.hidders_at_REMOVETHIS.pandora.be>
Date: Mon, 12 Jul 2004 19:02:55 GMT
Message-Id: <pan.2004.07.12.19.03.20.22495@REMOVETHIS.pandora.be>

On Mon, 12 Jul 2004 10:01:31 -0700, Neo wrote:

>> > Because the three strings each represent the same thing, the string
>> > 'brown', it is redundant.
>> 
>> So you keep on claiming, but the standard definition of redundancy in 
>> normalization theory says otherwise.

>
> The "standard" definition of redundanty you are referring to is a
> limited definition most applicable to a limited data model (RM).

No. It's the general definition that is applicable to all data models, even such wild data models as XML.  

>> > I realize the following are unusual examples, however a general data
>> > model can't (application above it can) have prejudices as to what
>> > updates are unusual. Suppose, the world is taken oven by Islam and
>> > they desire every string in a computer to be spelled backwards, thus
>> > 'brown' needs to be updated to 'nworb'.
>> 
>> That means the schema is changing and then all bets are off because the
>> notion of update anomaly is defined only with respect to a fixed
>> schema.

>
> While all bet may be off in RM because anomaly only applies to fixed
> schema, such is not the case in my data model where even schema is
> considered just another thing.

No. Also there all bets are off because you have to redefine what exactly an update anomaly is and motivate that definition.

• Jan Hidders

Received on Mon Jul 12 2004 - 14:02:55 CDT