The Fuller Stack
summary
Argues that “full stack” must include hardware, physics and people, not just front- and back-end frameworks, and classifies software into S/P/E classification (specifiable, partially specifiable like ML, and embedded systems that feed back into the world). Cites concrete failures: MasterCard test-format bugs, Facebook timestamp showing 1969, Intel 1970 updates, the 2038 32-bit time limit, a race car steering NaN failure, an art project using 99 phones to fake traffic, and warns CAP = choose two and that NFTs are receipts rather than ownership.
AI-written summary · about
chapters (13)
- 0:00The full stack concept
- 4:21Classifying code systems
- 7:43System interactions
- 11:09Low-level constraints
- 15:28Managing state and errors
- 19:40Physics and latency
- 22:06Limits of logic and data
- 25:35Measuring real value
- 28:25Roadmaps and planning
- 32:14Business value paradox
- 36:11Collective intelligence
- 41:46Scaling and distribution
- 45:04CAP theorem and conclusion
related talks
54mFull Stack Development: What Does It Really Mean?
08mFull Stack Developers: Myth or Reality
45mBe Fullstack, because everyone should
08mFull-Stack Testing Ethos For a Progressive Team
41mWhy Full-Stack Is the Future of Web Application Development
35mCraft, Engineering, and the Essence of Programming
47mThe Meaning of LIVE
01h 09mThe Future of Software Engineering