Re: Re: Reusing abbreviated keys during second pass of ordered [set] aggregates

On Tue, Dec 22, 2015 at 10:49 PM, Peter Geoghegan <pg(at)heroku(dot)com> wrote:
> I attach a rebased patch for 9.6 only.

I marked the patch -- my own patch -- "Ready for Committer". I'm the
third person to have marked the patch "Ready for Committer", even
though no committer bounced the patch back for review by the reviewer,
Andreas:

https://commitfest.postgresql.org/9/300/

First Andreas did so, then Michael, and finally myself. The problem is
that you see things like this:

"Closed in commitfest 2015-11 with status: Moved to next CF"

Michael (the CF manager at the time) remembered to change the status
to "Ready for Committer" again; you see this entry immediately
afterwards:

"New status: Ready for Committer"

but I gather from the CF app history that Alvaro (the current CF
manager) did not remember to do that second step when he later moved
the patch to the "next" (current) CF. Or maybe he just wasn't aware of
this quirk of the CF app.

I don't have a problem with having to resubmit a patch to the next CF
manually, but it's easy to miss that the status has been changed from
"Ready for Committer" to "Needs Review". I don't think anyone ever
argued for that, and this patch was accidentally in "Needs Review"
state for about 5 days. Can we fix it?

--
Peter Geoghegan