Building Consistently Fast Flutter Apps
summary
Shows concrete checks and numbers: measures frame build time with Flutter DevTools and the Performance Overlay in profile mode, targets 16 ms per frame but sees spikes up to 33 ms on the raster thread. Demonstrates using integration_test APIs (traceAction and watchPerformance) to capture baselines, exports Chrome tracing and summary CSVs, and automates regressions in CI (Firebase Test Lab, cloud device farms). Recommends testing critical flows, isolating heavy work with isolates, and failing CI on persistent regressions.
AI-written summary · about
chapters (11)
- 0:00Performance testing in Flutter
- 3:16Premature optimization flaw
- 7:14Real devices and profile mode
- 10:03Performance overlay for jank
- 13:12Flame charts for bottlenecks
- 15:59Performance testing loop
- 18:27Integration tests for performance
- 21:41Interpreting test results
- 26:03Automate performance checks in CI
- 27:30Speed versus deadlines
- 30:21Key takeaways
related talks
49mDemystifying performance tune up in Flutter and Dart
38mDebugging performance issues with the Flutter DevTools
40mFluttering at 60 FPS
41mThe art of Flutter testing
38mFlipping the Testing Pyramid
44mNo More Jank: How to Keep Your Flutter App Smooth and Snappy
32mChasing performance issues methodically
28mChasing 60 FPS: Frame Time Profiling in Jetpack Compose