Taking Ember Offline
summary
An open source Ember app called HospitalRun aims to run in places with flaky Internet and proves offline web apps work in production—deployed in 2013 for medical data collection. It uses Application Cache then moves toward Service Workers, stores offline data with PouchDB + CouchDB synchronization, and solves conflicts by per-field timestamps. Lessons: partition client sync to avoid heavy downloads, throttle sync for old machines, and expect search and conflict-handling tradeoffs when syncing rich datasets.
AI-written summary · about
chapters (23)
- 0:00Intro
- 0:44Background
- 2:17Offline Web Apps
- 3:37Hospital Run
- 4:59Why the Internet is Being Attacked
- 6:23Offline App Server Downtime
- 7:33Offline First Approach
- 8:33Offline First Website
- 8:55Stop treating offline as an error
- 9:16Develop a survivor mentality
- 10:03Application Cache
- 11:05Ember CLI App
- 11:38Demo
- 13:27Offline Storage
- 14:30Offline Data Persistence
- 15:12CouchDB
- 16:55Synchronization
- 17:56Synchronization Patterns
- 19:27Application Cache vs Service Workers
- 21:12Service Workers in Ember
- 21:32Broccoli Service Worker
- 21:58Ember Service Worker
- 23:58Questions
