Summary
The article details how Grab successfully implemented R8 optimization in its Android app to tackle performance challenges, particularly high Application Not Responding (ANR) rates. It describes three core challenges encountered: slow debugging due to obfuscated code, testing at scale with a large codebase, and slow feedback loops. The team addressed these with innovative strategies including AI-assisted debugging, a pragmatic testing approach that identifies and resolves common patterns, and an optimized feedback loop through remote compilation. The results include significant improvements in app size, startup time, and stability, highlighting the feasibility of R8 optimization even at a large scale.