On Mon, 14 May 2001, Andrew Snow wrote:
> I note that it still is not possible (v7.1.1) to have a table with a foreign
> key check on an inherited column for rows in child tables. Will this ever
> be supported, and what is the status of work on this feature?
It will eventually be supported. There's a bit of support work that needs
to come first (inherited unique constraints) and a fair amount of work on
the fk stuff itself. I think that getting the fk to completely work on
normal tables (there are still some problem cases with deferred
constraints and the constraints really should be referencing
oids and shouldn't use a self-conflicting lock for locking rows) probably
happens first.