Thomas Lockhart writes:
> Extra pairs of eyes are helpful here; can anyone see that TIME is
> excluded from the types defined for OVERLAPS (which would free us to Do
> It Our Way) or if the spec calls for an implementation different from
> the part of the spec I found (which might be The Right Way)?
No, the current implementation is correct.
The drawback with redefining the time data type to be a circular number
line is that it leads to definitional problems in other areas of the
arithmetic. For example, what would the result of
time '3:00' - time '23:00'
have to be?
A wrapping time type would probably be useful, but not when it shadows the
standard type.
--
Peter Eisentraut peter_e@gmx.net