Performance in a Kotlin World
summary
Shows how to find real Android bottlenecks using Android Studio profilers — CPU profiler, memory, energy and network — and system tracing (cystrace/Perfetto) plus the Jetpack Benchmarking library for CI. Example: schedule screen with ~200 sessions spends seconds in eager LiveData+map/filter chains that allocate multiple lists and block rendering; startup traces reveal Firebase init and Dagger graph work on application onCreate. Fixes include deferring menu/tags, converting eager transforms to a lazy thunk, and adding targeted benchmarks to detect regressions. cystrace
AI-written summary · about
chapters (9)
- 0:00Introduction to performance
- 4:03Why performance matters
- 10:04Profiling Android apps
- 16:45Kotlin performance myths
- 22:25RecyclerView flame charts
- 28:50Optimizing Kotlin code
- 32:01Benchmarking in CI
- 38:01Advanced tracing tools
- 42:59Summary and tools
related talks
45mPerformance in a Kotlin world
47mPerformance in a Kotlin World
43mBenchmarking your application
35mIn Search of Performance: the Subtle Art of Android Profiling
37mImproving app performance with benchmarking
45mHigh performance in Android
32mChasing performance issues methodically
39mPerformance for Conversion! Use Distributed Tracing to Pinpoint Bottlenecks