Evidence Over Advice: Learning Programming Faster Using the Scientific Method
summary
Argues for treating programming claims like experiments: makes a concrete claim about useMemo, runs microbenchmarks with 10,000+ iterations, and reports a surprising 90% reduction on heavy sorts while noting trivial gains become nanoseconds; runs Node experiments that show promises run before setTimeout and explains microtasks; builds a web‑worker demo that removes main‑thread jank so UI stays responsive. If deciding on attendance, expect measured numbers, real code, and working demos rather than abstract advice.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 1:26Skepticism improves knowledge
- 6:02Scientific method for programming
- 8:50When to experiment
- 9:33Testing useMemo performance
- 13:25useMemo benchmark results
- 15:34Resolving confusion with microtasks
- 17:42Microtask experiment results
- 19:51Curiosity and web workers
- 21:53Web worker benefits demo
- 23:17Experimental mindset
- 24:22Q&A
related talks
31mHow to be better
41mHow to Get People Excited about Functional Programming
40mThe Best Programmer I Know
27mThe Many Ends of Programming
14mA Book Nerd's Guide to JavaScript
27mThe myth of the “Real JavaScript Developer”
52mKeynote: A speedrunning guide to software development
29mThe Undeniable Virtue of Doing Less