Re: modelling history in a database
From: Anton Versteeg <anton_versteeg_at_nnll.iibbmm.com>
Date: Fri, 14 Feb 2003 11:30:09 +0100
Message-ID: <3E4CC531.2506F59C_at_nnll.iibbmm.com>
Date: Fri, 14 Feb 2003 11:30:09 +0100
Message-ID: <3E4CC531.2506F59C_at_nnll.iibbmm.com>
Depending on the usage of the history table,
one could leave away the 'new data' part,
since the 'new data' can be found somewhere else:
either in a following history entry or in the original table.
stu wrote:
> "Jason" <pmorrison_at_4lpi.com> wrote in message
> news:cf188409.0302131022.3be0e1ad_at_posting.google.com...
> > Jan Marten Visser <jan.marten.visser_at_virgil.nl> wrote in message
> news:<3E4A2B18.C729DD29_at_virgil.nl>...
> > Second, you could create a table Change that has fields for the table,
> > field, old data, new data and date of change.
> >
>
> I like that idea! Have you tried it in practice? Does anybody else have
> any comments on this suggestion?
>
> Cheers
> Stu
-- Anton Versteeg DB2 Specialist IBM NetherlandsReceived on Fri Feb 14 2003 - 11:30:09 CET