At a glance
Jazzband is sunsetting. New signups are disabled. Project leads will be contacted before PyCon US 2026 to coordinate transfers. The wind-down plan has the timeline. The retrospective looks back at the past decade.
Over 10 years ago, I started Jazzband as a cooperative experiment to reduce the stress of maintaining Open Source software. Everyone who joined got access to push code, triage issues, and merge pull requests. “We are all part of this.”
It’s time to wind things down.
GitHub’s slopocalypse, the flood of AI-generated spam PRs and issues, has made Jazzband’s model of open membership and shared push access untenable.
Jazzband was designed for a world where the worst case was someone accidentally merging the wrong PR. Now only 1 in 10 AI-generated PRs meets project standards. Curl had to shut down its bug bounty after confirmation rates dropped below 5%. GitHub’s own response was a kill switch to disable pull requests entirely. I can’t keep giving push access to everyone who joins and expect to operate Jazzband safely.
The cracks have been showing for much longer, though.
Jazzband was always a one-roadie operation. People asked for more roadies and offered to help over the years. I tried a number of times to make it work, but it never stuck. I dropped the ball on organizing it properly. When volunteers stepped up, they’d step back after a while. I don’t blame them. Volunteer work needs a structure that supports people, and I hadn’t provided one.
Every project transfer, lead assignment, PyPI permission change and infrastructure decision still went through me.
The sustainability question was raised as early as 2017. Five years into Jazzband, I gave a keynote at DjangoCon Europe 2021 about it. I said out loud that the “social coding” experiment had failed to create an equitable community, and that there wasn’t a sustainable solution without serious financial support.
I presented a roadmap to revamp the infrastructure, grow the management team, formalize guidelines and reach out for funding. None of that happened. The PSF fiscal sponsorship was the one thing that did.
In the years since, I’ve been on the PSF board, which faced its own crises, and now serve as PSF chair. That work matters and I don’t regret prioritizing it. But Jazzband got even less of my time.
Meanwhile, GitHub moved in the opposite direction. Copilot launched in 2022, trained on open source code that maintainers were burning out maintaining for free. 60% of maintainers are still unpaid.
The XZ Utils backdoor in 2024 showed what can happen when a lone maintainer burns out and someone malicious fills the gap. Jazzband’s own infrastructure was also getting in the way. The release pipeline couldn’t support trusted publishing, and projects that needed admin access were stuck waiting.
Projects started leaving. That’s OK. Moving on was always supposed to be part of the deal.
I want to thank Django Commons and Tim Schilling for picking up where Jazzband fell short. They have 5 admins and 15 active projects, including django-debug-toolbar, django-simple-history and django-cookie-consent from Jazzband. django-polymorphic is transferring over right now. They solved the governance problem from day one. If you’re a Jazzband project lead looking for a new home for your Django project, start there.
For non-Django projects like pip-tools, contextlib2, geojson or tablib, I’m not aware of an equivalent. If someone wants to build one for the broader Python tooling ecosystem, I’d love to see it.
Over 10 years, Jazzband grew to 3,135 members from every continent but Antarctica, maintained 84 projects with ~93,000 GitHub stars, and shipped 1,312 releases to PyPI.
Projects that passed through Jazzband are downloaded over 150 million times a month. That includes pip-tools at 23 million and prettytable at 42 million. django-debug-toolbar spent 8 years under Jazzband and ended up in the official Django tutorial. django-avatar, a repo from 2008, was still getting releases in 2026. django-axes shipped 129 versions, with a release every 13 days in its peak year.
The full 10-year retrospective has more of the numbers and stories.
I’m not pulling the plug overnight. The detailed wind-down plan covers the timeline.
Timeline
If you’re a project lead, expect an email soon.
Thanks to the 81 project leads who kept things going despite the bottlenecks I created, and to everyone who joined, contributed, filed issues and shipped releases over the years. You were strangers on the internet who decided to maintain something together. Jazzband depended on that willingness to show up.
I started Jazzband because maintaining Open Source alone was exhausting. The irony of becoming a single point of failure for 71 projects is not lost on me. Even so, people collaborated, maintained projects and shipped releases. Those projects will move on to new homes, as they were always free to do.
We are all part of this.
Written by Jannis Leidel on Mar 14, 2026, 12:00:00 PM