The Imposter’s Guide To Dependency Injection
summary
Shows that constructor injection and a simple analytics tracker interface replace direct Firebase calls so unit tests can assert tracking without Android runtime crashes. Demonstrates a manual dependency container, then maps that recipe to Hilt vs Koin: Hilt uses compile-time modules and annotations, Koin registers modules at runtime. Covers scoping tradeoffs like application-scoped vs component-scoped instances and the build-time/runtime failure differences, and argues practical benefits: easier testing, swap‑ability of vendors, less boilerplate, and when to start manual versus adopting a library.
AI-written summary · about
chapters (14)
- 0:00Introduction to talk
- 1:49Imposter syndrome and community
- 2:30Code without dependency injection
- 4:39Constructor injection basics
- 6:45Should we mock Firebase Analytics?
- 8:10Creating custom interfaces
- 9:55Sharing dependencies across screens
- 12:45Manual dependency containers
- 13:48Hilt dependency injection
- 16:57Koin dependency injection
- 18:42Benefits of dependency injection
- 21:13Advanced concept of scoping
- 24:41Choosing a DI solution
- 29:56Q&A
