Sponsored Session: Why Don't You Love Me Back, Open-source Software?
summary
Moves from proof-of-concept to product require concrete actions: builds using Yocto/meta layers and QEMU then real hardware (example: Qualcomm RB5) reveal boot-time, memory and frame‑rate regressions that QEMU missed; teams end up with a growing “backpack” of divergent patches and expensive maintenance. The saved‑hours claim is false unless infrastructure exists: GitLab/GitHub autobuilders, device lab, SBOM tooling, CVE gating and policies, and a plan to upstream work (budget 20–25% developer time) to keep upgrades feasible.
AI-written summary · about
chapters (22)
- 0:00Open-source relationships intro
- 1:47Seven stages of open-source love
- 3:17Infatuation with projects
- 5:42Dating in virtual envs
- 7:28Physical dates with platforms
- 8:54Struggle in production
- 11:01Forking and patching
- 12:23Rebasing performance
- 14:55Disillusionment and debt
- 16:23Cost of ownership
- 19:07Stability phase
- 21:37Supply chain ingredients
- 25:30Investing in infra
- 28:39Compliance and usability
- 31:50Upstreaming debt
- 33:16Influencing projects
- 35:26Speed versus stability
- 36:45Q&A on analysis tools
- 38:14Q&A on business trade-offs
- 40:45Q&A on debt decisions
- 42:33Q&A on minimizing debt
- 43:18Closing remarks
related talks
43mSponsored Session: Why Don't You Love Me Back, Open-source Software?
01h 02mBoF: The Yocto Project and OpenEmbedded
31mFrom Closed To Collaborative: Perspectives and Lessons From Qualcomm’s Open Development Experience
13mJust how far can you go with FOSS?
36mEmbedded Development Made Easy: A Guide to FOSS Platforms and Maxing product value
50mDo our development tools need a refresh?
47mBoF: Yocto Project & OpenEmbedded
24mBaremetal Software ❤ FOSS