Designing Hypermedia APIs
summary
Claims that JSON alone can't drive hypermedia and that HTTP + proper MIME types + links must be the engine of application state. Walks through five concrete steps: evaluate business processes, model a state machine, derive media-type needs, define hypermedia controls, implement. Demonstrates a working example: a sitemap/webpage validation API using application/vnd.w3clove.validation+json with hypermedia controls (links and forms) that clients follow instead of hardcoding URLs, showing how representations guide workflows rather than RPC-style endpoints.
AI-written summary · about
chapters (9)
- 0:00Introduction and motivation
- 2:35Redefining REST as hypermedia
- 4:45Architectural styles of APIs
- 7:29Understanding hypermedia
- 10:25Core definitions and semantics
- 12:18Components of hypermedia APIs
- 15:33Practical design process
- 18:34Implementing a state machine
- 25:44Conclusion and resources
related talks
44mDesigning Hypermedia APIs
43mThe Hypermedia API
33mDesigning RESTful Web Services
33mHypermedia – the solution for future web applications?
20mYour Http-based API is not restful
58mHypermedia APIs and the Future of AI Agentic Systems
30mThe internet of things challange:building APIs that last for decades
43mHow the Web Evolves with Hypermedia