Re: Relational Database Design Practices: Deletion?

"Bernard Peek" <bap_at_shrdlu.com> wrote in message news:E7QCPMitFc8+EwDA_at_diamond9.demon.co.uk...
> In message <_FjIa.166$Dx4.24616052_at_mantis.golden.net>, Bob Badour
> <bbadour_at_golden.net> writes
> >"Adam Weiss" <adam_at_removespamfodder.signal11.com> wrote in message
> >news:3ef14b53$1_at_news.qnet.com...
>
> [...]
>
> >If you go to http://www.pgro.uk7.net/books.htm and search for "Temporal"
on
> >the page, you will find a book that describes the ideal way to handle
what
> >you are doing. From that, you should be able to adapt a solution that
works
> >around SQL's lack of interval type generators.
>
> I'm not certain that this is what Adam's problem is. It sounds to me as
> if the solution for him may be a data warehouse approach. That involves
> purging the live database of data that is probably obsolete while
> keeping a second copy in a separate database.

Apparently, you have never read the book. It deals explicitly with the issue of designing databases with current, future and historical data. It covers the logical difference between deleting only the current data and with deleting both the current and historical data. The first delete says that something was valid and is no longer value while the second delete says that something was never valid.

Basically, it boils down to the difference between a "delete" and a "using delete". Received on Thu Jun 19 2003 - 16:33:44 CEST