Speeding up Startup for Large React Apps
summary
Claims concrete wins: machine learning–driven packaging reduces redundant downloads sitewide, server-side module aliases and JS calls resolve branches and stream initial data so clients fetch only needed code, a bootloader API and matchRoute abstraction let UI parts load asynchronously, and an inline requires transform plus requestIdleCallback scheduling cuts execution before render by around 10% and makes JS unit tests almost twice as fast — these systems are built and producing measurable improvements.
AI-written summary · about
chapters (12)
- 0:00Start
- 4:44Branches in Code
- 12:17Can the server do more?
- 13:45Streaming initial data with JS::call
- 17:25Minimizing Initial JS Bytes
- 18:19Bootloader
- 20:07Route based branching
- 21:15matchRoute
- 23:26CommonJS
- 24:57Inline Requires
- 26:30Lazy Execution: Trade Offs
- 28:44Recap
related talks
25mDesigning High Performance React Applications
25mThe first two seconds: Faster page loads with React
26mOptimizing React Native Performance
29mBuilding Large-Scale Web Apps
34mOptimizing Performance in Large-Scale React & Node.js Applications
33mUniversal React Apps Using Next.js
13mSpeed at Scale: Optimizing The Largest CX Platform Out There
27mBuilding High Performance React Applications