Re: To Denormalise or not to Denormalise

From: Paul Druker <pdruker_at_metaway.com>
Date: 2000/04/26
Message-ID: <zmqN4.57507$cZ.96207@typhoon.southeast.rr.com>#1/1

Lawrence,

Actually, as soon as you finish modelling your "well designed" database, you do the last step -- you denormalze it to achieve better performance.

Paul

"Lawrence" <lsimela_at_mahalini.prestel.co.uk> wrote in message news:8e58ks$4lh$1_at_phys-ma.sol.co.uk...
> Hi All
>
> I wonder if any of you folk out there can help settle an argument with a
> colleague who passionately believes that one should never have to
> denormalise any part of a well designed database. He believes that a
'well
> designed' database should be in 3NF or higher and can be efficiently
> implemented without denormalising any part of it.
>
> Thanks in advance for any comments
>
> Lawrence
>
>
Received on Wed Apr 26 2000 - 00:00:00 CDT