OSS vs. Open Source Software: What's Different? What's the Same?
summary
Running engineering for IBM's Accelerated Discovery Platform shows open-source science prioritizes reproducibility of results and treats models and datasets as the crown jewels, not just code; licenses (GPL, Apache, MIT) and model terms (Llama research vs commercial) shape adoption. Practical takeaways: use Git and automated CI for versioned assets and tests, manage library versions and large data outside GitHub, and leverage science’s decade‑scale experience in distributed high‑performance compute.
AI-written summary · about
chapters (10)
- 0:00Introduction and background
- 2:42Defining open source and open science
- 8:22History of openness in software
- 11:33Motivations for open source
- 14:43Motivations for open science
- 17:30Lessons from open source for science
- 20:59Reproducibility and testing
- 24:09Language ecosystems and dependencies
- 27:17Lessons from science for software
- 28:39Summary and call to action
related talks
42mOpen Source at a Crossroads
21mOpen Source Software and the Next Decade of Innovation
30mNo OSS, No Users: Why Investing in Open Source Matters
15mOpen Core vs Open Source
57mWhat You Need to Know About Open Source—Trust Me, I'm a Lawyer
33mThe Cathedral, the Bazaar, and the Coffee House
37mOpen-Source Science: Working with OSPOs to Accelerate Research in...
34mConsumers To Contributors: Open Source As a Competitive Advantage