Robert Haas <robertmhaas@gmail.com> writes:
> Now, if six people who are all well-known PostgreSQL contributors show
> up and they all say "I looked at the latest version of this carefully
> and I'm confident you've got it right", then go ahead: push it. But
> don't make the mistake of thinking that because you're confident that
> you've now got it right everybody else will like it too. Even since
> you posted the last version, Heikki expressed a concern that resulted
> in (surprise!) more revisions. There comes a point where a patch that
> is still heavily in flux is just too late for the release cycle, and
> we're well past that point at this stage of the game.
FWIW, I agree that we're past the point where we should be committing
features whose external definition hasn't been stable for awhile. Fixing
bugs post-feature-freeze is one thing, but if there's a significant chance
that you'll be having to adjust the feature definition, then it's probably
too late for 9.5. And this particular item sure looks like it's in that
category.
There's always another release cycle.
regards, tom lane