Lightning Talk: Resolvers Everywhere: Rethinking Client and Server Boundaries in GraphQL
summary
Argues that GraphQL execution can live on clients and demonstrates Meta’s Relay Resolvers: client-defined fields that fetch an encrypted blob via a fragment, decrypt into a decrypted note type, and cache values; resolver functions run lazily today and can be live resolvers with read/subscribe for reactive UI. Tradeoffs: fast to add and kept out of long‑lived server schema, but must be implemented per client, costs device CPU for heavy work, and Relay’s approach isn’t a full GraphQL executor.
AI-written summary · about
related talks
25mDemystifying GraphQL Clients
58mApolloClient or Relay with Fragments, ‘Hairy’ GraphQL and TypeScript
14mWhy Relay is a Must for your GraphQL APIs
32mGraphQL-Anywhere - How GraphQL can improve your work across all of your stack
32mReintroducing Relay
43mGraphQL in the Era of Server Components
20mHow GraphQL Servers Break and How to Harden Your Resolvers
31mFrom Relay to Apollo, a story of GraphQL clients