Re: Modelling objects with variable number of properties in an RDBMS

From: FrankHamersley <>
Date: Wed, 02 Nov 2005 11:33:12 GMT
Message-ID: <Yt1af.6984$>

VC wrote:
> "Jonathan Leffler" <> wrote in message
> news:YHY9f.4683$

>>VC wrote:
>>>"Roy Hann" <specially_at_processed.almost.meat> wrote in message 
>>>>"vc" <> wrote in message
>>>>>Roy Hann wrote:
>>>>>>Actually no, let's not.  What you have re-invented here is the tired 
>>>>>>entity-attribute-value (EAV) design (much beloved of medical research
>>>>>>wonks for some reason).
>>>>>Often they do that (EAV) not because they do not know any better but in
>>>>>order to overcome a specific database physical limitations. Their
>>>>>entities very often contain thousands of attributes (molecular
>>>>>biology/pharmaceutical research/drug discovery) [snip].
>>>>Always check the bath water for babies...
>>>I beg your pardon ?
>>There's an old saying "Don't throw the baby out with the bath water".
>>In this context, it means (roughly) make sure you don't lose the benefits 
>>of the relational model by thowing away all the rules that help

> I know that, but what recipe does Roy suggest for, let's say, MS SQL Server
> that does not have either UDTs or other means to implement an entity with a
> higher than the table limit number of attributes, beyond offering a cute
> saying ?

It seems axiomatic to me if you have reached this point in your requirements you have selected the wrong database.

Was MS SQL selected before the requirements were expressed (1 black mark) or after (10 black marks)?

Cheers, Frank. Received on Wed Nov 02 2005 - 12:33:12 CET

Original text of this message