On Tue, May 18, 2021 at 7:44 AM Justin Pryzby <pry...@telsasoft.com> wrote: > As things stand, in this case I think it *should* be included, since the > backpatched change isn't the same as the change to HEAD (removing the GUC). > The git_changelog output might well be wrong in this case (or, arguably, the > "remove the GUC entirely" should've been a separate master-only commit than > the > "make the GUC do nothing" commit).
I suppose that's true -- maybe it should be listed separately, because the GUC is removed in 14 only. > However, Peter indicated an intent to add a reloption to disable the vacuum > optimization, so maybe the removal of the GUC could be documented at that > time. This is unrelated to the bypass indexes in VACUUM thing. -- Peter Geoghegan