Re: Commitfest Update - Mailing list pgsql-hackers

On 7/15/22 14:57, Justin Pryzby wrote:
> On Fri, Jul 08, 2022 at 02:41:52PM -0700, Jacob Champion wrote:
>
>> If there are no objections, I'll start doing that during next Friday's
>> patch sweep.
> 
> I think it's fine to update the cfapp fields to reflect reality...
> 
> ..but a couple updates that I just saw seem wrong.

Hm, okay. Let me hold off on continuing then; I'm only about 25% in. The
general rule I was applying was "if you were marked Reviewer prior to
June, and you haven't interacted with the patchset this commitfest, I've
removed you."

>  The reviewers field was
> nullified, even though the patches haven't been updated in a long time.
> There's nothing new to review.  All this has done is lost information that
> someone else (me, in this case) went to the bother of adding.

My understanding from upthread was that we wanted to get out of the
habit of using Reviewers as a historical record, and move towards using
it as a marker of current activity. As Tom said, "people see that the
patch already has a reviewer and look for something else to do."

I am sorry that I ended up reverting your work, though.

> Also, cfapp has a page for "patches where you are the author", but the cfbot
> doesn't,

(I assume you mean "reviewer"?)

> and I think people probably look at cfbot more than the cfapp itself.

I think some people do. But the number of dead/non-applicable patches
that need manual reminders suggests to me that maybe it's not an
overwhelming majority of people.

> So being marked as a reviewer is not very visible even to oneself.
> But, one of the cfbot patches I sent to Thomas would change that.  Each user's
> page would *also* show patches where they're a reviewer ("Needs review -
> Reviewer").  That maybe provides an incentive to 1) help maintain the patch; or
> otherwise 2) remove oneself.
I didn't notice cfbot's user pages until this CF, so it wouldn't have
been an effective incentive for me, at least.

Also, I would like to see us fold cfbot output into the official CF,
rather than do the opposite.

> Also, TBH, this seems to create a lot of busywork.

Well, yes, but only because it's not automated. I don't think that's a
good reason not to do it, but it is a good reason not to make a person
do it.

>  I'd prefer to see someone
> pick one of the patches that hasn't seen a review in 6 (or 16) months, and send
> out their most critical review and recommend it be closed, or send an updated
> patch with their own fixes as an 0099 patch.

That would be cool, but who is "someone"? There have been many, many
statements about the amount of CF cruft that needs to be removed. Seems
like the CFM is in a decent position to actually do it.

--Jacob