Six Impossible Things
summary
Six concrete claims and counterexamples: floating point NaN and infinity cause real failures (a MATLAB-controlled racecar steers into a wall), a classic binary search overflow bug lurks at 2^31-1, divide-by-zero can cripple a ship, permutation/bogo sorts show nontermination so timeouts are how systems cope, epoch/time limits (1970/2038) break assumptions, and technical debt is a metaphor not a euro amount. Expect examples, standards (IEEE 754, POSIX), demos and practical takeaways about testing and roadmaps.
AI-written summary · about
chapters (17)
- 0:00Six impossible things
- 1:36Infinite representations
- 4:15Floating point myths
- 6:23Correctness and bugs
- 10:38Nature of assumptions
- 13:05Limits of execution
- 13:27Time as a resource
- 17:04Algorithm limitations
- 22:06The halting problem
- 24:31Undecidable propositions
- 30:45Uninitialized data
- 33:59Engagement vs data
- 35:54The unknowable future
- 38:12Roadmaps and value
- 44:55Distributed systems
- 50:27Technical debt metaphor
- 56:41Q&A and conclusion
related talks
01h 05mTalk Six Impossible Things
02h 04m6 impossible things
54mSeeing Like a Programmer - Resiliency, Limits, And Moral Hazards In SE
45mProgramming's Greatest Mistakes
17mIf You’re Happy and You Know It (Inside the Mind of a Developer)
01h 02mZen and the Art of Code Lifecycle Maintenance
40mUnaccustomed As I Am to Public Speaking
01h 28mWhat Do You Mean?