Storing app data reliably & performantly with SimpleStore
summary
Claims SharedPreferences silently fails and blocks UI; at global scale millions of devices make a 99.9% success rate still mean thousands of failures and P95 storage latency ~20ms. A new library SimpleStore implements fully async IO with a thread pool and ordered executors per namespace, exposes listenable-future callbacks (optional Rx adapter), stores bytes not XML, supports proto support, is closeable to release caches, and is open-sourced on GitHub after a soft launch showing smoother UI and explicit failure callbacks.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:58Storage reliability at scale
- 9:15Storage performance matters
- 13:24API ergonomics for storage
- 16:36Introducing SimpleStore
- 21:42Design principles of SimpleStore
- 28:05Configuring SimpleStore
- 31:19Storing complex objects
- 32:47Future of SimpleStore
related talks
27mSharing my preference for DataStore
30mDataStore Preferences and migrating from SharedPreferences
19mJetpack DataStore - protocol buffers storage solution
25mThe New Modern Data Persistence on Android - Meet Jetpack Datastore
38mThe Dangers of SharedPreferences: An API allegory
24mLightweight Realtime & Offline Data Storage on Android without a Database
34mHow to reactively load and cache data without even trying
21mStore Library