Re: A Normalization Question
Date: 12 Jul 2004 09:50:56 -0700
Message-ID: <4b45d3ad.0407120850.36f52327_at_posting.google.com>
> I understand the reason not to store the string 'brown' more than once,
"A small step for man, a giant leap for mankind..." :)
> but how do you store the fact that Mr. Brown lives on Brown St.,
> drives a brown car, has brown hair, wears a brown suit,
> and has one brown and one blue eye?
By using a data model more flexible than RM. While it is not impossible to do in RM, it requires generic modelling (ie all data in a few tables) and is impractical. I will post the above data with an alpha version of XDb2 in a few days.
A poor analogy: RM is like a fixed set of wrenches (ie 1mm to 10mm in certain increments) that covers the most commonly used sizes. Therefore it is fast/efficient. In comparison, XDb2 is like one variable wrench, it takes more steps to set up for a particular size but that process can be used from 0mm to 20mm. For most general apps, RM is the correct choice. For some apps, such as AI apps, RM falls short. Received on Mon Jul 12 2004 - 18:50:56 CEST