Failure is not an option...
summary
Recounts concrete engineering choices from Apollo 11: the Apollo guidance computer with ~1 MHz CPU, core rope memory woven by seamstresses, and 420,837 lines of code (25% comments, ~11,000 pages) that ran on a few kilobytes, handled repeated 1201 alarms during descent and landed with 17 seconds of fuel. Shows practices that worked: end-to-end simulators, asynchronous priority scheduling, strict change flow resembling source control branching, and rigorous fault-protection that prevented catastrophic failures.
AI-written summary · about
chapters (37)
- 0:00Introduction
- 0:49Neil Armstrong on the Moon
- 1:32Context
- 2:40The Moon
- 3:05Apollo Guidance Computer
- 3:32JavaScript
- 3:57Margaret Hamilton
- 4:29Software Engineering
- 5:40Engineer
- 6:06KCD
- 6:44Hamilton
- 7:07True Engineering
- 7:32End to End Testing
- 9:20Core Rope Memory
- 9:45Shipping to Production
- 10:11Testing
- 11:11Apollo 8 Earthrise
- 12:18Apollo 11 Launch
- 13:02Lunar Landing
- 14:36Mission Control
- 14:48Hamiltons Reaction
- 15:17Hamiltons References
- 15:51Agile
- 16:01Scrum
- 16:37Source Control Management
- 17:17Apollo Guidance
- 17:44What can we learn
- 18:02Dont assume
- 18:17Astronauts
- 18:36Lego
- 18:46The key
- 19:06The wrong Margaret Hamilton
- 19:27Standing on the shoulders of giants
- 19:46Thank Margaret Hamilton
- 19:59Engineering Mindset
- 20:36Thank you
- 20:50One more thing
related talks
27mExplosive overflow: Lessons from rocket science
51mFailing Successfully: Reporting and Handling Errors
01h 01mPeople that make computers go crazy
19mHow Do You Even Write Secure Code Anyways
50mThe cost of code
36mKeynote: The Future of Software 
36mMove fast Build things… Safely!
22mWhat Could Go Wrong? The Subtle Science Of Coding for Failure