Why React Is Becoming the Next jQuery
React is already the next legacy framework. Most teams just haven't started dreading it yet.
Framework wars aren't settled on Hacker News. They're settled by hiring pipelines, bootcamp curricula, and what junior engineers learn first — which is exactly how React won in 2013 and why that win is now a liability. Eleven years of surface area, a Server Components rollout that took years to stabilize, and docs that contradict themselves by version: the cracks are visible if you're willing to look.
Svelte compiles away the runtime entirely. HTMX went from meme to production at companies that know exactly what they're doing. The migration pressure is real. But React has a new moat nobody's pricing in: LLMs were trained on vastly more React code than any other framework, which means AI tooling defaults to React, which means new codebases default to React, which means the next generation of training data is React. The moat isn't community anymore.
The reversal is in the architecture. React won when JavaScript was the only serious option. The next era has WebAssembly, edge runtimes, and server components that dissolve the client-server line so aggressively that the framework abstraction underneath it may not survive. The winner by 2028 is probably already in your node_modules. You've been ignoring it because the docs weren't polished yet. That's how the last three transitions happened.
Chapters
0:00 The Claim
0:33 React's Unchallenged Decade
1:01 The Cracks Appearing
1:26 The Challenger: Svelte
1:50 The Real Signal: HTMX
2:17 The Distribution Problem
2:30 The AI Variable
2:59 The Reversal
3:27 The Bet, Not the Best
Links
Svelte — https://svelte.dev
SvelteKit — https://kit.svelte.dev
HTMX — https://htmx.org
Vercel — https://vercel.com
React Server Components RFC — https://github.com/reactjs/rfcs/blob/...
State of JS Survey — https://stateofjs.com
GitHub Copilot — https://github.com/features/copilot