A real-life journey into the opinionated world of 'utility-first' CSS
summary
Argues that switching to utility-first CSS stops most custom stylesheet churn and speeds work dramatically, showing a real refactor from a messy BEM codebase to tachyons and a small flexbox-grid where roughly fifty custom modifiers disappear, the site rebuilds in ten days while camping, CSS drops from ~275KB to less than half, and a new blog ships with no custom CSS, using simple utilities like mt0 to solve recurring layout tweaks.
AI-written summary · about
chapters (13)
- 0:00Start
- 1:46Separation of CONCERNS
- 7:06CSS gets hard to UPDATE
- 9:18The best way to maintain CSS is to stop writing CSS.
- 13:10"Un-semantic" classes are breaking the web.
- 13:21Confusion between semantic HTML and semantic classes.
- 15:29You stop writing CSS!
- 16:18Ridiculously fast WORKFLOW No, seriously!
- 19:27What about specificity? Design restrictions?
- 21:48CHRISTMAS BREAK
- 25:06Stage 1: Tachyons as SUGAR on top.
- 27:29UTILITY. LAST versus UTILITY FIRST
- 31:08Copy & paste mess
