x wrote:
> I would add that your interval is a set and you know the relational model
> don't like sets as attributes.
There is a certain truth in this proposition. If we consider an
interval as a set of points, then the set of all intervals covering a
given point (or, more generally, a set of points) is formally a
relational division. Relational division is a prominent example of set
join. Of course, the fact that an interval is an infinite set is very
inconvenient from technical perspective.
Received on Fri Feb 10 2006 - 14:23:35 CST
