>>>>> "Bruce" == Bruce Momjian <bruce(at)momjian(dot)us> writes:
Bruce> I think there is general agreement that we should have a
Bruce> timezone data type which validates against
Bruce> pg_timezone_names().name.
What happens when pg_timezone_names output changes? (which it can do,
especially if the install is using the OS tzdata; even if not using OS
tzdata, it's not expected to be stable even between point releases)
--
Andrew.