I want to clarify a couple things that might convince the people opposed
to this (and if not I have a compromise proposal). Closing a PR doesn't
mean we lose the potential contribution. You can still find closed PRs
here:
https://github.com/sagemath/sage/pulls?q=is%3Apr+is%3Aclosed+is%3Aunmerged. I
think rather than thinking of closing a PR as deleting it, it's more
accurate to think of it as archiving the PR. With my proposal there
would also be a tag to filter to the abandoned ones. I think this would
provide an easy way to search for abandoned PRs to take over and have
someone else complete.
There are two reasons I go to https://github.com/sagemath/sage/pulls:
1. To find PRs to review myself.
2. To get a sneak peak for what new features/fixes might appear in an
upcoming version of Sage.
A PR that was reviewed with changes requested and no developer response
doesn't fall into either of the two categories above, which is why I
think these abandoned PRs are mostly noise.
But if people are still opposed to moving the PR from the open tab to
the closed tab, instead of closing it we could just add a tag for
"unresponsive" but still leave those PRs open. We could then say that
any PR marked with the unresponsive tag is fair game for someone else to
take over (preferably after asking the original author if they intend to
finish it).
Adding an "unresponsive" tag (either in the open or closed tab of pull
requests) would communicate that you shouldn't expect that feature to be
added anytime soon, but if you want to do that work yourself someone has
already started it and you can take over.
I think functionally leaving it open or closed doesn't change much, it's
the "unresponsive" tag that makes it clear that someone else is free to
take the PR over. I would prefer to close abandoned PRs just to clean up
the list of open pull requests, but if people prefer to leave it as open
and just have the "unresponsive" tag then I'd be okay with that as a
compromise.
To provide some additional context, the PR that prompted me to post this
was this: https://github.com/sagemath/sage/pull/38292, it adds a new
feature (Enigma encryption) that's only really of pedagogical/historical
interest and hence low-priority. Unless there's a Sage contributor with
both the interest and knowledge in the Enigma cipher and the time to fix
up the PR, it's unlikely that it will get picked up by someone else.
There are at least a dozen other encryption schemes that would be more
useful to add to Sage, so while interesting it's a pretty niche
contribution that I doubt anyone will want to pick up and finish since
most of the people who have the expertise to finish it would probably
want to work on adding an encryption scheme that is of interest to
current cryptography researchers. Despite multiple people working on the
original PR, there has been no response from the developers since I
reviewed it in July. As somewhere between my original proposal and my
compromise proposal, we could add an "unresponsive" tag and leave it up
to discretion whether the PR is important enough to leave open or if
it's better of closing it because it's unlikely to get picked up. Then a
search for the "unresponsive" tag on open PRs would return a list of
abandoned PRs that add/fix something important, and a search for the
"unresponsive" tag on closed PRs would return a list of abandoned PRs
that while they would be cool to have, aren't really important for Sage.
Vincent Macri (He/Him)
--
You received this message because you are subscribed to the Google Groups
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to sage-devel+unsubscr...@googlegroups.com.
To view this discussion visit
https://groups.google.com/d/msgid/sage-devel/b3569a20-a9fb-4f27-83e5-766918353811%40ucalgary.ca.
