API Development in 2017
summary
Company moves from Jbuilder to ActiveModelSerializers and then to schema-first development using OpenAPI (Swagger); engineers write OpenAPI, serve mock servers, and auto-generate clients, validators and server stubs with Swagger UI/Editor/Codegen and a community validator. Team extends OpenAPI: default-required fields, an optional: true flag, external $ref and a merge feature, plus auto-generating serializers from schemas (handling dates, AR types). Notes: Grape used earlier, GraphQL adoption grows, and a BFF layer appears for client-specific aggregation; releases run about three APIs per year.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 1:04Web API history and REST
- 6:06Resource design principles
- 9:49JSON generation evolution
- 13:07JSON Schema introduction
- 16:21API description language trends
- 17:28Schema-first development
- 21:58Schema challenges and improvements
- 28:33GraphQL benefits
- 33:51GraphQL vs REST comparison
- 37:27BFF layer usage
- 37:54Summary and future outlook
related talks
19mSecure and accelerate APIs' lifecycle with the OpenAPI specification
43mAPI first with Swagger and Scala
35mAPI-First! Contract-First! Make API development great again!
27mBuilding Great Web APIs
30mAPI Code First vs Design First
40mOpenAPI Generator: 4000 stars on GitHub and spaghetti
16mAPI-First Development with OpenAPI — You Should Practice It!?
27mGetting developers hooked on your API