A Beginners Guide to Open Source Metrics and Analysis
summary
Data about open source is often inconsistent and changing: GitHub Archive queries show inconsistent counts (same time window yields different totals), bot activity skews metrics — an estimated 120 million bot events in 2020 and 44% of PR events from bots in 2020 with growth from ~3% (2017) to 44% (2020). Practical fixes include stating exact definitions, sources and queries, respecting privacy and licenses, and using tooling like Grimoire Labs, Augur, CNCF DevStats.
AI-written summary · about
chapters (21)
- 0:00Intro
- 0:25Sophias background
- 0:49Data is messy
- 1:02Data issues
- 2:03Ambiguity
- 3:08Source Code Logs
- 4:26Inconsistencies
- 4:56Bots
- 6:15Bots on Github
- 6:59People
- 7:22Layers
- 8:19Accountability
- 9:03PI
- 10:27Data Sharing
- 11:54Policies Licenses
- 12:50Policy and Regulation
- 14:17Privacy Documentation
- 15:41Good Data Practices
- 17:16Misleading Metrics
- 19:51Tools
- 21:56Conclusion
related talks
35mRegulations, Ethics And Messy Data: Tales From An Open Source Analyst
35mThe Benefits and Pitfalls of OSS Project Metrics: Measuring Health and Risk in Open Source Communities
30mData Tales from an Open Source Research Team
08mAll eyes on Open Source: Monitoring community sovereignty
43mBuilding and Supporting Open Source Communities Through Metrics
45mWhat’s Going On In A FOSS Project?
40mOpen Source Is Not the Same Anymore
26mBuilding & Supporting Open Source Communities Through Metrics