Panel: Fostering better collaboration between the scientific Python community and core Python...
summary
Claims anchor around concrete wins and pain points: work on the NumPy array API and the matrix‑multiply operator improves interoperability, faster CPython-style incubators let experiments land before wider adoption, and packaging efforts like Wheel Next plus CUDA containers tangibly cut install pain. Data shows surprising scale — JetBrains reports 48% in data roles — and forums are failing: discuss.python.org creates toil. Expect proposals, repo names, benchmarks, install counts, and whether prototypes actually run on Windows and GPUs.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 1:31Panelist backgrounds
- 3:15Bridging scientific and core Python
- 9:41Scientific Python community
- 12:52Why collaboration matters
- 17:26Key bridging projects
- 23:09Incubator programs and CPython
- 28:25Openness versus scalability
- 32:39What works and what doesn’t in collaboration
- 34:24Discourse and PEP challenges
- 39:22Audience takeaways and closing thoughts
- 44:16Final audience questions and responses
related talks
55mAccelerated Python: The Community and Ecosystem
54mKey Challenges In the PyData Ecosystem and How we Can All Make a Difference
47mBig ideas shaping scientific Python: the quest for performance and usability
44mKeynote
36mLightning Talks
01h 00mKeynote: Python now and in the future
58mPython Data Ecosystem: Thoughts on Building for the Future
38mKeynote: Python for Pythonistas