The trends in choosing licenses in Python ecosystem
summary
A forensic scan of top 2,500 PyPI packages uses the PyPI JSON API and a Python script to extract declared licenses and finds BSD most common, MIT and GPL families and Apache around 12% each, and ~6% unknown/no-name entries; a surprisingly large fraction lack a clear license. The project generates pull requests to add/clarify licenses and reports ~20 PRs merged so far, plus concrete recommendations: ship full license text, copyright headers, setup.py classifiers, and use Apache for patent concerns.
AI-written summary · about
chapters (11)
- 0:00Start
- 8:50UNKNOWN
- 9:40Software License
- 10:30Commercial
- 11:37Polic Domain Tublic Domain
- 13:06Open Source
- 16:13Reciprocal Licenses
- 17:16Choose a proper license based on usecase
- 17:48LICENSE File in your source code
- 19:29Do not invent your own license
- 20:03Keep your funny bones aside
related talks
35mOn the structure and reproducibility of Python packages - data crunch
24mBuilding a Sustainable Python Package Index
01h 04mLightning Talks
32mPyPI in the face: running jokes that PyPI download stats can play on you
27mPSF Welcome
35mSecuring Python projects Supply Chain
28mWhyPy? Understanding Trends Driving the Usage of Python
52mKeynote - Ee Durbin