Comparing API Protocols - One Feature at a Time
summary
Compares GraphQL, REST, OpenAPI and gRPC feature-by-feature and shows concrete tooling: Hive (Gateway, registry, Edge CDN), GraphQL Mesh and a CLI composition library that converts OpenAPI/gRPC/SOAP into GraphQL, and the acquired State cache running on Fastly with partial query caching implemented in Rust. Claims GraphQL lets services see exactly which fields are used, enabling safer evolution and deprecation and argues that 85–90% of APIs don't follow their OpenAPI spec, so GraphQL’s runtime schema and tools (Pothos, codegen) solve real operational problems.
AI-written summary · about
chapters (8)
- 0:00Introduction and experiment
- 5:18Recap of recent announcements
- 10:35Brainstorming API features together
- 18:48Service-to-service communication debate
- 27:44When to avoid GraphQL
- 32:22Simplicity vs complexity trade-offs
- 37:19Domain-specific API standards
- 41:22Closing thoughts and call to action
related talks
01h 15mREST, GraphQL, and gRPC: A Comparison
01h 00mThe Secrets to Smart, Flexible, Self Describing API
59mGraphQL, gRPC or REST? Resolving the API Developer's Dilemma
17mBoost API Performance: gRPC vs REST vs Async
48mHow to Design APIs: Swagger, gRPC, GraphQL — Which to Choose?
47mA tale of query languages. Is GraphQL The Chosen One?
30mGraphQL, gRPC and REST, Oh My! A Method for Unified API Design
43mTUTORIAL: We've "REST"ed long enough -- let's have an API adventure!