A Possible Future of Embedded Software Development
summary
Argues that embedded needs concrete fixes not slogans and demonstrates working pieces: an RTFM kernel that uses interrupt hardware as a priority lock to eliminate deadlocks and bound interrupt latency, a template metaprogramming pipeline (Conan + SVD-derived generators) that emits efficient drivers that beat vendor C, and a proof‑of‑concept bounded buddy allocator that gives an upper time bound for allocations. Claims static exceptions and freestanding C++ toolchains can drop 40 KB footguns and that these pieces scale toward thousands not millions of lines.
AI-written summary · about
chapters (15)
- 0:00Intro
- 1:44The Problem in Bare Metal
- 2:34Hard RealTime Component
- 5:01What is Hard RealTime
- 6:52Locality of Reasoning
- 14:11Interrupt Service Routine
- 16:03Atomics
- 17:30An awesome person
- 20:22Mythic language
- 22:33Standard library
- 24:15Crossplatform
- 25:57UART Driver
- 27:17Code Generator
- 29:22Sleep
- 31:09Conclusion
related talks
30mThe Era of Vibe Coding: Why High-Skill Engineers Are More Critical Than Ever
18mBuilding pi in a World of Slop
43mThe Future of Programming
44mJoint Keynote Address: Accelerating Software Development: The LLM (R)evolution
01h 09mThe Future of Software Engineering
28mWhat Handcrafted Servers Taught Us About Handcrafted Code- 44mAccelerating Software Development: The LLM (R)evolution
36mEngineering Intent: Solving the AI Productivity Paradox