Stephen Frost <sfrost@snowman.net> writes:
> * Tom Lane (tgl@sss.pgh.pa.us) wrote:
>> Yeah, I can reproduce this. I suspect it got broken by Stephen's hacking
>> around with default ACLs. A simple example is
> Yes, it's related to the work I did with pg_dump's ACL handling, because
> we're no longer just always including the whole revoke/grant set of ACLs
> for everything in the output.
I see that this is listed on the open items for v11, but it's hard to
justify it being there, because the bug exists only in 9.6 and 10.
(We fixed it in HEAD as part of the pg_dump-vs-pg_dumpall refactoring.)
I gather, from the lack of any progress since November, that you're
probably not going to fix it in the back branches. I'm not excited
about working on it either, but I dislike leaving such a bug unfixed.
In any case, I think it should be removed from the v11 list. The "older
bugs" section is meant to capture pre-existing bugs that we might possibly
fix as part of v11 stabilization, and this isn't that.
regards, tom lane