Rethinking Syncing
summary
Claims that replacing REST sync logic with a real‑time NoSQL stack (Couchbase + sync server + Couchbase Lite on iOS, alternatives: Realm, Firebase) cuts most syncing code: model objects use Swift Codable, replication runs continuous push/pull replicators, and a demo app with two devices keeps lists instantly in sync, so it works. Caveats: large documents slow sync, need channel design and migration strategy, no built‑in NSFetchedResultsController equivalent, and you can run multiple replicators to control priorities.
AI-written summary · about
chapters (18)
- 0:00Intro
- 1:33Syncing Variables
- 3:45To-Device Syncing Approaches
- 5:34The Net Change Approach
- 11:10From-Device Syncing Approaches
- 13:21Complications to REST API Syncing
- 15:16REST API Syncing Summary
- 17:43The idea
- 21:56What just happened
- 23:13How do we implement?
- 23:39Model Objects
- 24:50Data Controller cont
- 26:51Fetching Data for Display
- 29:27Handling Refresh on Change
- 29:55Adding Data
- 30:44Deleting Data
- 33:59Eventually consistent DB Disadvantages
- 37:35Eventually consistent DB Best Use Cases
related talks
41mSync is the Future of Mobile Data
47mThe future of Real-time/offline/data
23mWhat's an App Without Data?
24mDon't be afraid to handle network loss in your code
27mNetworking & Persistence in Mobile Applications at Scale
35mSync Data with your Mobile Kotlin App The Easy Way!
40mBoost App Reliability: Offline Storage Solutions with AWS Expertise
19mReactive iOS App Development