On 7/15/22 16:42, Jacob Champion wrote:
> If you have thoughts/comments on this approach, please share them!
Okay, plenty of feedback to sift through here.
[CFM hat]
First of all: mea culpa. I unilaterally made a change that I had assumed
would be uncontroversial; it clearly was not, and I interrupted the flow
of the CF for people when my goal was to be mostly invisible this month.
(My single email to a single thread saying "any objections?" is, in
retrospect, not nearly enough reach or mandate to have made this
change.) Big thank you to Justin for seeing it happen and speaking up
immediately.
Here is a rough summary of opinions that have been shared so far; pulled
from the other thread [1] as well:
There are at least three major use cases for the Reviewer field at the
moment.
1) As a new reviewer, find a patch that needs help moving forward.
2) As a committer, give credit to people who moved the patch forward.
3) As an established reviewer, keep track of patches "in flight."
I had never realized the third case existed. To those of you who I've
interrupted by modifying your checklist without permission, I'm sorry. I
see that several of you have already added yourselves back, which is
great; I will try to find the CF update stream that has been alluded to
elsewhere and see if I can restore the original Reviewers lists that I
nulled out on Friday.
It was suggested that we track historical reviewers and current reviews
separately from each other, to handle both cases 1 and 2.
There appears to be a need for people to be able to consider a patch
"blocked" pending some action, so that further review cycles aren't
burned on it. Some people use Waiting on Author for that, but others use
WoA as soon as an email is sent. The two cases have similarities but, to
me at least, aren't the same and may be working at cross purposes.
It is is apparently possible to pull one of your closed patches from a
prior commitfest into the new one, but you have to set it back to Needs
Review first. I plan to work on a CF patch to streamline that, if
someone does not beat me to it.
Okay, I think those are the broad strokes. I will put my [dev hat] on
now and respond more granularly to threads, with stronger opinions.
Thanks,
--Jacob
[1]
https://www.postgresql.org/message-id/flat/34b32cb2-a728-090a-00d5-067305874174%40timescale.com#3247e661b219f8736ae418c9b5452d63
