Rx Concurrency: A Prescription for Multi-threading
summary
Compares AsyncTask pitfalls to RxJava by showing concrete fixes: replacing 18 lines with four lines of Rx keeps parsing logic unchanged, moves network work off UI by using subscribeOn vs observeOn so tests complete in ~7–8 ms while retries run later, and demonstrates an exponential backoff implemented with map/flatMap/delay on the IO scheduler is an unbounded thread pool. Lifecycle leaks are handled with Disposable or ViewModel; solution separates threads, reduces shared mutable state, and works when cancellation is cooperative.
AI-written summary · about
chapters (8)
- 0:00Introduction and setup
- 1:23Problems with AsyncTask
- 9:26Introducing RxJava basics
- 12:34Understanding subscribeOn and observeOn
- 21:51RxJava vs RxAndroid
- 26:12RxJava schedulers explained
- 33:43Handling memory leaks in Rx
- 37:42Architecture and best practices
