The month-old PR

This is getting frustrating. In short, --enable-libgdbm-compat is enabled for every known gdbm distribution except Homebrew, and was needed for a project of mine. I figured it would be a simple PR, given the precedent and scope.

That was over a month ago. Since then, there have been 10 replies. Two of which asked questions which I had already answered (it’s enabled on all other distributions, and CI is failing because it builds all reverse dependencies and some of those are failing for unrelated reasons), and the other 8 replies have been me, providing status updates and trying to get the attention of people.

IMHO, the larger issue seems to be triage. Formulae bumps have always been handled quickly and easily for me (thanks!), because they are low hanging fruit. But beyond that, it gets progressively harder to convince someone with merge access to carve out time for a thorough review, and the PR is effectively dead if it’s dropped out of the first page (after bumps are processed).

The docs address this, kind of, but only with:

Post a comment on your pull request if you’ve provided all the requested changes/information and it hasn’t been merged after a week.

But that’s really not working here. Some sort of active “I believe there are no blockers and that replies have been addressed” assertion would be nice; basically the flip side to the “maintainer feedback” tag. (I wish Github allowed for a class of submitter-specified tags.)

I don’t have hard data here, but I imagine it’s happening for more PRs than mine, ones which get lost to “stale” because the submitter gave up. I was considering it myself, but wanted to post this instead to see if something good can come out of it myself.