"Kevin Grittner" <Kevin.Grittner@wicourts.gov> wrote:
> Tuple locks should be safe from that because we use the tuple xmin
> as part of the target key, but page and heap locks
That should have read "page and relation locks".
> I guess that tips the scales in favor of it being worth the extra
> code. I think it's still in that gray area
I just thought of something which takes it out of the gray area for
me: pg_locks. Even though it would be extremely rare for a false
positive to actually occur if we let this go, people would be sure
to get confused by seeing locks on the dropped objects in the
pg_locks view.. They've got to be cleaned up.
-Kevin