Lightning talk from Adam Wolff
summary
Proposes building a reactive, live-updating GraphQL server around an immutable update log with entries ordered by epoch time plus randomness, append-only rows for nodes and edges, and readers that read backwards to get the latest state. Adds a memoize over topics API to recompute values only when underlying topic iterables change, supports partial writes and eventual consistency, and implements streaming tail load for connections. Ships a toy Newsfeed class prototype (a bit cumbersome) rather than benchmarks or production metrics.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 0:50Live updates
- 1:45Data structure
- 3:00Streams
- 3:18Topic
- 3:38Memo Eyes
- 5:05Partial Rights
- 5:44Connections
related talks
58mCrafting compelling real-time web experiences with GraphQL and React
09mLightning Talk: Making GraphQL Fun for the Backend Too
30mReact state management in a GraphQL era
34mGraphQL real-time subscriptions
32mGraphQL Future
27mLightning Talks May 4th
23mGraphQL in real life
24mPanel discussion hosted by the Apollo team on GraphQL + React