On Wed, Dec 13, 2017 at 1:46 AM, Peter Eisentraut <peter.eisentr...@2ndquadrant.com> wrote: > OK, here is a bigger patch set that addresses these issues. I have > added OBJECT_RELATION to reflect the difference between TABLE and > RELATION. I have also renamed NAMESPACE to SCHEMA. And then I got rid > of AclObjectKind as well, because it's just another enum for the same thing. > > This is now a bit bigger, so I'll put it in the commit fest for detailed > review.
Patch 0001 is simply removing EventTriggerSupportsGrantObjectType(), but shouldn't we keep it and return an error for objects that have no GRANT support? Returning conditionally true looks like a trap waiting to take someone in.. I mentioned that in https://www.postgresql.org/message-id/cab7npqsr8rsh-rcmjv7_2d7oksbyre4frhueyn_krehgo_y...@mail.gmail.com already, but this has remained unanswered. Similarly, not using default in the switches of stringify_adefprivs_objtype() and stringify_grantobjtype() would be nice to grab warnings during compilation. And patch 0002 is doing it the correct way in aclcheck_error(). -- Michael
