conda-forge: supporting the growth of the v...
summary
conda-forge runs as one-recipe-per-repo with 17,000 repositories and about 3,800 maintainers, producing nearly 300 million downloads per month; the index data (repo_data.json) swells to ~120 MB and CDN sync shows a roughly two-hour delay. automation and bots manage feedstock review, stage recipes gate new packages, and tooling efforts use mamba/libmamba for faster SAT solving, zchunk and OCI mirrors for smaller repo downloads, and workarounds like repodata patching to avoid full rebuilds.
AI-written summary · about
chapters (20)
- 0:00Intro
- 0:25We are part of the conda-forge core team
- 1:58Historical context for conda
- 4:50Channel maintenance in pre-conda-forge times
- 5:53Emergence of conda-forge
- 6:46Foundational principles
- 8:12Immediate success!
- 9:26Key benefits
- 11:51Who runs conda-forge
- 12:37How does it work?
- 18:06Azure, GitHub Actions, cirun.io
- 20:08Failovers: more mirrors
- 23:41Mamba CZI grant: more improvements coming!
- 25:21mambabuild & boa
- 25:50Continuing the growth
- 27:00Technical challenges and improvements
- 29:49Supporting mamba from conda
- 32:00Scaling for the community needs
- 33:45Governance and Open Source
- 36:26How can you help?
related talks
37mConda Forge - Community Driven Packaging That Works for You
34mRising Threat: Securing Your Open-Source Pipeline
30mBringing Continuous Delivery to Open Source
05mQuantifying your reliance on Open Source software
17mBuilding Secure Open Source Communities From the Ground Up
01h 08mChallenges of Managing CoreFX repo
58mBuilding an Open Platform for Sustaining Data Science Innovation
20mAutomating maintainers work: The Prototype