Code Generation for Zombies
summary
A CSV-to-source pipeline reads workflow screen event columns and emits concrete Java classes (Analytics, Workflow, Screen, event methods) plus a pluggable tracker interface so providers swap without code churn. It uses JavaPoet to generate code, uses J2ObjC to target iOS, and claims 100% accuracy while reducing integration time and increasing developer productivity. The analytics tracker generator yields autocomplete, compiler-checked removals, smaller manual testing surface, and cross-platform runtimes that work in production.
AI-written summary · about
chapters (26)
- 0:00Intro
- 3:10Dashboard
- 4:34Android
- 6:10Data Management
- 6:55Problem Solving
- 7:56Cognitive Resources
- 9:29How do we solve this
- 11:07Idea Relation
- 13:39Objective
- 14:11Blueprints
- 16:24Analytics
- 19:30Analytics Class
- 20:02Workflow Class
- 20:25Inner Classes
- 21:02Design
- 21:20Swift
- 22:04C
- 22:24Swift Compatibility
- 22:53Objective C
- 25:24ID Feature
- 26:26Autocomplete
- 26:46implicit checklist
- 27:07testing
- 29:03cogeneration tips
- 30:42tools
- 32:50questions
related talks
48mMobile Developer Productivity at Scale
39mDespaghettifying the hairball
43mAndroid Dev Tools and Workflows
40mAutomating Android Workflows with Github Actions
43mDoppl Code Sharing
37mGenerating Kotlin Code
47mDealing With KSP in Common Code for Android and iOS
47mClaude Code at Scale: An Android Deep Dive