Peter Eisentraut <peter_e@gmx.net> writes:
> Tom Lane writes:
>> Yeah. We had that discussion at some point during the 7.3 development
>> cycle, and concluded we liked table-local naming for constraints better
>> than the SQL spec's global constraint names.
> Then why does pg_constraint store a schema OID?
I think we left it there for possible future use with assertions, which
are not tied to any particular table.
regards, tom lane
