"Greg Sabino Mullane" <greg(at)turnstep(dot)com> writes:
>>> SELECT *,has_schema_privilege(oid,'USAGE') FROM pg_namespace;
>> Well, if you test it as a superuser, it's going to return TRUE every
>> time.
> Exactly. So I'm not seeing how we can use USAGE as a reliable test for
> the case where a temporary table was created by the same user, but in
> another session.
Superusers can access anything they want to. What's your point?
The spec says "accessible" ...
regards, tom lane