Is the gunicorn project dead?

[tnusser]

Looking at the frequency of commits/changes and reviews I have the feeling, that not much is happening anymore. The project website seems to be not updated at all anymore. There are 98 open PRs which wait on being reviewed etc. For the sake of future-proofing our tech stack, is it advisable to still use gunicorn for new projects?

This is by no means any criticism to the maintainers of this project as they created and maintained this the last decade probably on a non-profit basis. I know this is a ungrateful job to do and am deeply thankful to what this project has become. But still I'd like to know if we can expect regular updates in the future in regards to security, bugfixes and new features.

[pajod]

In any case, it will be kept alive (if not by the hand of existing maintainers, then in additional forks of big users), for as long as a couple of other projects require something like Gunicorn, yet so far none of the drop-in-alternatives are a clear upgrade path for all-around better API design, documentation, portability, correctness, upgrade stability and performance. For the most part, Gunicorn does what it was meant to do, and does require little regular change. By my count only 10 serious bugs really need to go into a general release right now, other bug reports are minor corrections or new features, or useful to keep Gunicorn interoperable & easy to maintain for the longer term. Apart from that, there are many overlapping/duplicates in the current PRs right now, yet a whole bunch of bulk-closed (due to age) issues still clearly relevant.

I offer my help with whatever way forward seems appropriate, specifically in the following areas:

• If you have changes that you want to see tested & reviewed, point me to them.
• I maintain personal issue/PR prioritization lists that I could use to triage GitHub issues and PRs, so it is easier for new contributors or merge decisions to keep track. (I do not wield permissions in this repo, but I tested earlier how I can share the same thing using GitHub projects or meta-issues with lists of links. Will do one of those if requested.)
• I do maintain a private fork without EoL compatibility, and with consistent code formatting and (by now, almost complete) type annotations added. I still plan to put that into something easily installable, either as a type standalone checking stub package or as a full package with inline annotations. (benoitc has indicated no interest in including them in this project at this time)
• Since I run a version with multiple PRs applied anyway, I could prepare something like this preview again, so people willing to test outstanding pull requests can apply something consistent and report back in bulk that they are happy with many PRs at the same time, then to be merged in bulk. (benoitc has indicated that merging entirely independent PRs one by one is still preferred)

[benoitc]

Gunicorn isn't dead—it's just moving slowly lately! The project could benefit from smaller, targeted patches to keep things streamlined.

I've been a bit sidetracked recently, but I'm closely following discussions. Comments from @pajod especially resonate with me. We have some duplicate or off-topic PRs to sort through, and there are several important ones I'll be merging soon.

Good news: A new version of Gunicorn will land this month, bringing numerous fixes and refocusing the project on its core strengths. Initially, it'll be available on a dedicated branch for review and thorough testing.

Additionally, I'll soon publish the refreshed website along with updated documentation. The new contribution guide should make the review process smoother and more collaborative.

Thanks everyone for your patience and continued support!