Changing human behavior for better code
summary
Argues that changing code requires changing people, not just policies, and gives a repeatable justify / understand / plan / execute / analyze / track process. Implements a custom ESLint/stylelint rule and a reporting script that aggregates violations by team; initial enforcement in CI triggers a revert when a team ships a big feature, leaving ~40 violations, then a second rollout adds editor plugins, office hours, training and documentation and drives violations to zero. Claims 50% of change initiatives fail and shows concrete metrics and tooling.
AI-written summary · about
chapters (5)
- 0:00Why code changes fail
- 1:08Change management process
- 5:06JUPET framework steps
- 13:17Real-world example
- 19:18Key takeaways summary
