To begin with, I鈥檒l explain: I respect the developers of Pipenv and am very glad that such tools are being developed.
But damn it: for more than a year, no releases, no answers, no roadmaps - nothing. Chat in slack is also empty. And this is with the officially recommended dependency management tool!
Okay, two months ago we received a response from the maintainer @techalchemy, he promised a rollout soon and hung a critical label on the release task:
https://github.com/pypa/pipenv/issues/4058#issuecomment-565550646
Hurrah?
Not so far. From July last year and to the present moment, he hasn鈥檛 done a single commit (correct me, if I'm wrong) in this repo.
And on the pulse, it generally seems that the project is dead. For comparison, look at the pulse of poetry
The last commit to the master was made in November 2019 and breaks the tests, but who cares!
Okay, I鈥檒l write again that my goal is not to offend anyone. But at the moment we have the main tool of the language, which is not supported by the maintainer. Which, in turn, do not want to somehow share their plans or solve a problem.
It's time to face the truth, look into your issues: you have failed completely to support the most important project for the python community. Admit this officially, and not in a hidden issue, but on the main page and write about your solution.
As I see the solutions (from the softest to the most radical):
355 issues and 32 pull requests seem a bit much for a one or two hobbyist main maintainers. Seems like pipenv needs funding to get somewhere again. I just contacted the Python Software Foundation by mail on this matter.
@hartwork Please write here when you receive an answer from the PSF, thanks!
@hartwork Please write here when you receive an answer from the PSF, thanks!
I got one reply now. While feel that it's not appropriate for me to quote that reply verbatim here, it seems fair to say that I was pointed to the PSF grants program and to work with the current maintainers of the project. That makes good sense to me.
Based on the Git history on master
of last 12 months, I obtained this list of people:
# git log --since="1 year ago" master | fgrep Author | sed -e 's,frostming,Frost Ming,' -e 's,[email protected],[email protected],' -e 's, <.*>,,' | sort | uniq -c | sort -n -r | sed -e 's,Author: ,,' | head -n5
305 Dan Ryan
89 Frost Ming
24 Tapasweni Pathak
11 John Vandenberg
10 Kenneth Reitz
I'll try contact these people now, putting you @lucidyan in CC.
Is there a way for people to donate to this project? I couldn't find a link in the docs
@hartwork any updates?
@hartwork any updates?
I got a first short reply from @techalchemy on 2020-02-22, replied back to him on 2020-02-23 with questions, waiting for a reply since then. No replies from the other four people so far.
@techalchemy
@frostming
@tapaswenipathak
@jayvdb
@kennethreitz
Will you give us any comments on the status of the project?
I spoke with @techalchemy in IRC today and pipenv now has an update in https://github.com/pypa/pipenv/issues/3369#issuecomment-594875477 .
Thanks to everyone who commented in this issue.
Pipenv is working toward a new release -- see #3369 for specific details. The goal is to get that out this month.
I helped @techalchemy write a new update email he's just posted on distutils-sig (mirrored on the pypa-dev list). Dan notes:
As some of you may already know, I recently changed jobs and am now working for Canonical where I have part of my time allocated to open source work in general (and pipenv in particular).
So: thanks, Canonical!
Dan's email also includes a few ways you can help, directly and indirectly, to expedite this release.
- Write a clear roadmap, make it public domain and attract people to contributing
And - where you wrote "Choose a new Maintainer" - I'd say:
- Add new maintainers
As a step towards a clearer future roadmap: @frostming @uranusjr @techalchemy I see that this milestone was closed in mid-2018 but isn't quite finished. Should those issues go into a milestone for, say, a July release?
Some of them would be doable, e.g. Activate PIPENV_VENV_IN_PROJECT with --local. But both the Pipfile override and lock file metadata issues seem to be very involved and too far from being implementable IMO.
Per Dan's announcement just now the project's going through some process, communication, & related changes. @techalchemy after you get through the next release, I think working with @frostming and @uranusjr on a roadmap would be a good way to organize what would go into the subsequent release.
This roadmap and contribution process issue is, as I understand it, where folks will talk about improving the release cadence in the future. @techalchemy @frostming and @uranusjr are in charge of this, but I think next steps are:
and I believe this sequence will go faster or slower depending on what Pipenv finds out in Step 2 (how much money and time is available even to do this). Could take a month, could take a year. Personally, I have donated maybe 16 hours to helping shepherd the most recent release and making comments like this one, in case that is a useful data point. (I don't plan to donate more time to Pipenv this year.)
Is there a way for people to donate to this project? I couldn't find a link in the docs
I'll quote Dan's post to the Discourse forum:
If your company uses Pipenv, ask them to sponsor Python packaging via the PSF so there鈥檚 a stronger fund for packaging work.
Tidelift has an estimated $371.12/month available for the maintainers ("lifters").
Most helpful comment
355 issues and 32 pull requests seem a bit much for a one or two hobbyist main maintainers. Seems like pipenv needs funding to get somewhere again. I just contacted the Python Software Foundation by mail on this matter.