Re: Does Codd's view of a relational database differ from that ofDate & Darwin? [M.Gittens]
From: Ged Byrne <ged.byrne_at_gmail.com>
Date: 9 Jun 2005 02:32:27 -0700
Message-ID: <1118309547.280221.113760_at_g14g2000cwa.googlegroups.com>
Date: 9 Jun 2005 02:32:27 -0700
Message-ID: <1118309547.280221.113760_at_g14g2000cwa.googlegroups.com>
Alex,
The special values could be modelled in terms of there operators. This would define how they behave in joins and selections.
For example, the UNKNOWN value would cause an aggregation to return an UNKNOWN value, because not all of the information available to answer the question is available.
The user would have to explicitly filter out unknown values to ket a sum of the known values.
NON_APPLICABLE, on the other hand, wold not affect any aggregations, since it has no effect on the result.
Special values could then be definable within the database, reflecting the needs of the situation being modelled. Received on Thu Jun 09 2005 - 11:32:27 CEST