Re: relational reasoning -- why two tables and not one?
From: paul c <toledobythesea_at_oohay.ac>
Date: Thu, 15 Oct 2009 02:31:46 GMT
Message-ID: <mQvBm.48113$Db2.9339_at_edtnps83>
> Even then, it is historical data. For the same donor, all the
> information might be the same from one donation to the next, or it might
> all be different. Assuming one wants to record the information as it was
> (or as it was reported by the donor) at the time of the donation, few
> if any functional dependencies are likely to exist.
>
> Unless, of course, one creates a temporal database, but even then, the
> charity seems to collect no information about intervening points of time
> so a temporal database seems inappropriate too.
Date: Thu, 15 Oct 2009 02:31:46 GMT
Message-ID: <mQvBm.48113$Db2.9339_at_edtnps83>
Bob Badour wrote:
>...
> Even then, it is historical data. For the same donor, all the
> information might be the same from one donation to the next, or it might
> all be different. Assuming one wants to record the information as it was
> (or as it was reported by the donor) at the time of the donation, few
> if any functional dependencies are likely to exist.
>
> Unless, of course, one creates a temporal database, but even then, the
> charity seems to collect no information about intervening points of time
> so a temporal database seems inappropriate too.
I presume 'historical' means rows are 'written once', normally not replaced. If so that would be a good word to use if it is in fact one of the requirements. Received on Thu Oct 15 2009 - 04:31:46 CEST