Re: Normalization Question
From: pc <toledobythesea_at_oohay.moc>
Date: Fri, 21 Jan 2005 23:35:44 GMT
Message-ID: <klgId.15681$Qb.299_at_edtnps89>
>
>
> It is such a shame that you have to guess at future cardinality right up
> front when designing a relational structure, eh? Wouldn't it be great if we
> could design at a level where a change in multiplicity, like a change in
> size, type, or constraints, could be handled without a changing the logical
> model of the data? smiles. --dawn
>
Date: Fri, 21 Jan 2005 23:35:44 GMT
Message-ID: <klgId.15681$Qb.299_at_edtnps89>
Dawn M. Wolthuis wrote:
> "Leandro Guimarães Faria Corsetti Dutra" <leandro_at_dutra.fastmail.fm> wrote
> in message news:pan.2005.01.21.06.55.35.625887_at_dutra.fastmail.fm...
>
>>Em Fri, 21 Jan 2005 00:19:31 +0100, BG escreveu: >> >> >>>To abide by the rules of normalization, should the phone & date >>>fields be placed into their own tables? >> >>No, but that may be good design anyway as you may want to save >>more than one telephone of each type, for example.
>
>
> It is such a shame that you have to guess at future cardinality right up
> front when designing a relational structure, eh? Wouldn't it be great if we
> could design at a level where a change in multiplicity, like a change in
> size, type, or constraints, could be handled without a changing the logical
> model of the data? smiles. --dawn
>
in that case, what would be the point of the model?
p Received on Sat Jan 22 2005 - 00:35:44 CET