Re: Does Codd's view of a relational database differ from that of Date & Darwin? [M.Gittens]

> Anything that can be modeled with [nulls]
> can also be modeled without them, and to the extent that they
> are convenient this is mostly due to the fact that the possibilities for
> user-defined domains were too restricted.

That statement surprises me. What about outer join?

In application code I often run in to situations where there is either a legal value or else there is a special case, and good practice is *not* to use a special value, but rather to "protect" the value with another variable. This leads to the unfortunate fact that you still sometimes have a (meaningless) value in the primary variable.

Ah! But pehaps when you say user-defined domains were too restricted, you meant that they left off union types:

data MyParticularSituation = IHaveOne of int   | ElseIDont;

I think that covers the "special values" situation pretty well, but it seems overly manual for capturing the cardinality=0 case.

Marshall Received on Sat Jun 04 2005 - 16:54:42 CEST