Is InnerSource Commons good for open source?
summary
Using ecosyste.ms data (14M packages, ~280M repos, 25B dependencies) the team maps 108 publicly listed InnerSource Commons organizations to GitHub, builds a 500MB SQLite corpus and finds 3,776 repos publishing ~14,000 packages with ~8.25M dependent repos and ~29M downloads. About 10.6% of identified maintainers make ~4.6M contributions to ~100,000 OSS repos, roughly 4% of commits to dependencies and 1.2% of work on ~10K critical projects; Microsoft strongly skews totals. Code and the dataset are published and a free tool surfaces fragile single-maintainer or stale projects for targeted support.
AI-written summary · about
chapters (8)
- 0:00Introduction and background
- 2:37Defining InnerSource Commons in data
- 7:38Analyzing ISC members' open source impact
- 11:31Free as in puppies analogy
- 17:25Assessing contribution quality and gaps
- 19:50Identifying unhealthy dependencies
- 23:02Public data and ecosystem insights
- 25:48Q&A and final remarks
related talks
46mOpen Source, InnerSource - Similarities, Differences, Connections
30mInnersource: open source collaboration patterns beyond public FOSS projects
25mLooking at the Future of InnerSource
39mSOS: Sustainable Open Source
32mExemplars, Laggards, and Hoarders: A Data-driven Look at Open Source Software Supply Chains
13mHow can we truly arrive at "open-source everywhere"?
44m7 Things you can Learn from Microsoft's Open Source Experience
29mTop 5 Myths of InnerSource