Adopting Jetpack Compose @ Scale
summary
Large apps describe concrete wins and pain points: Airbnb started experimenting in April 2020, open sourced showcase, and ported a Compose design system over three years; Uber rebuilt its design system, runs on Bazel/IntelliJ and fought RecyclerView/interop performance issues that improved across Compose 1.1–1.5; Pinterest keeps Compose limited because heavy recycler feeds and years of view-level micro‑optimizations made performance gaps unacceptable; Reddit declares Compose‑first since January, enforces lint rules, uses snapshot tests and a CI page performance score to guard regressions.
AI-written summary · about
chapters (19)
- 0:00Welcome and introduction
- 1:02Panelist introductions
- 2:04Starting Compose adoption journey
- 7:42Training teams on Compose
- 13:43Early API design challenges
- 16:08Recommended learning resources
- 18:01Architecture with Compose
- 19:25Performance considerations
- 23:04Performance metrics and tracking
- 29:22Testing strategies for Compose
- 36:27Tooling and Android Studio
- 41:13Kotlin Multiplatform thoughts
- 44:56Design systems in Compose
- 51:47Accessibility in Compose
- 57:47Globalization and localization
- 1:02:01Widgets and wear OS
- 1:03:27Navigation approaches
- 1:05:15Experimental APIs policy
- 1:08:04Closing remarks and thanks
related talks
18mCompose Adoption at Pinterest: A Postmortem
34mRetrofitting Legacy Code with Jetpack Compose at Scale
37mAdopting Jetpack Compose Safely
42mBranching out to Jetpack Compose
17mWhat's new in Jetpack Compose
21mPerformance best practices for Jetpack Compose
43mBranching out to Jetpack Compose
37mSomething old, something new: Adding Jetpack Compose to a large open source Android app