Re: Storing derived and derivable data

From: Bob Badour <bbadour_at_pei.sympatico.ca>
Date: Mon, 01 May 2006 13:53:14 GMT
Message-ID: <epo5g.1348$A26.36813@ursa-nb00s0.nbnet.nb.ca>

Jan Hidders wrote:

> dawn wrote:
> 

>>Jan Hidders wrote:
>>
>>>dawn wrote:
>>>
>>Given that
>>the RM has as a goal not to have duplicate data and this is clearly a
>>duplication of data, I was thinking this might be one place where the
>>application would have the knowledge, rather than the DBMS.

Jan, how could you let such astonishing ignorance stand unchallenged? The RM says nothing about the physical duplication of data and actively encourages multiple logical views on the same data.

> Another goal of the RM is to shield the data from sloppy applications,
> so if duplication cannot be avoided then at least the DBMS should be in
> charge of it and not the applications.

Absolutely, I agree. Similarly for pointers. If one must have pointers to achieve adequate physical performance, the dbms should manage them to ensure correctness and to ease the burden on others. Received on Mon May 01 2006 - 08:53:14 CDT