6 impossible things
summary
Six precise traps decide whether a talk is worth watching: a reproducible claim like IEEE‑754 subtraction is functionally complete or the concrete fix midpoint prevents overflow; a demo or bug (binary search broken since 1983, a ship crippled by divide‑by‑zero, a race car stopped by NaN steering); an experiment that built something and failed (permutation/Bogo sort timeouts, ChatGPT hallucinating nonexistent APIs); and quantified system limits — CAP choices trade consistency availability partition‑tolerance — not vague promises about topics.
AI-written summary · about
chapters (28)
- 0:00Introduction
- 1:02Origin of six impossible things
- 3:11Representations cannot be infinite
- 7:27Floating point numbers aren't real
- 13:12Binary search bug that lasted decades
- 18:54Nature of assumptions in software
- 21:41Yorktown ship and divide by zero
- 24:11Memory management and undefined behavior
- 27:03ChatGPT and illusion of answers
- 28:27Problem with time in software
- 35:35Arithmetic mean of empty ranges
- 38:47Surprising behavior of NaN values
- 44:06Illusion of lost data
- 45:30Importance of understanding state
- 49:53Difference between null and empty
- 56:22Two hard problems in computer science
- 57:05Misnomer of algorithms library
- 1:01:47Permutation sort and algorithm illusions
- 1:13:44Halting problem and timeouts
- 1:16:57Kurt Gödel and undecidable propositions
- 1:21:31Myth of undefined behavior
- 1:24:43Five orders of ignorance
- 1:27:29Branch prediction and surprise costs
- 1:30:40Problem with product road maps
- 1:35:34Illusion of distributed systems
- 1:39:05Brewer's CAP theorem
- 1:44:20Truth about technical debt
- 1:55:14Conclusion
related talks
01h 07mSix Impossible Things
01h 05mTalk Six Impossible Things
40mUnaccustomed As I Am to Public Speaking
01h 32mKeynote: It Depends..
54mSeeing Like a Programmer - Resiliency, Limits, And Moral Hazards In SE
17mIf You’re Happy and You Know It (Inside the Mind of a Developer)
53mHow Not to Be Wrong in Software Engineering
01h 02mZen and the Art of Code Lifecycle Maintenance