Re: A Normalization Question
From: Jan Hidders <jan.hidders_at_REMOVETHIS.pandora.be>
Date: Tue, 13 Jul 2004 17:02:41 GMT
Message-Id: <pan.2004.07.13.17.03.05.533718_at_REMOVETHIS.pandora.be>
Date: Tue, 13 Jul 2004 17:02:41 GMT
Message-Id: <pan.2004.07.13.17.03.05.533718_at_REMOVETHIS.pandora.be>
On Mon, 12 Jul 2004 20:23:28 -0700, Neo wrote:
>> Sure, you can represent it that way, even in the relational model, but
>> you cannot call it normalization because normalization deals with removing
>> logical redundancies and there were no logical redundancies in the
>> original relation to begin with. I know you keep on denying this but
>> that's what the definition of logical redundancy says.
> > You may be correct by your limited definitions, but having the same > thing (ie a person, color, street, string or symbol) multiple times in > one db is redundant.
At the logical level that is an uninteresting type of redundancy. What matters there is if there are are update anomalies, and there are none iff there is no logical redundancy.
- Jan Hidders