A no nonsense GraphQL and REST comparison
summary
Argues that GraphQL is a focused query layer optimized for network speed and tight client–server coupling (born from Facebook’s FQL), while REST favors longevity, decoupling and HTTP features like sparse field sets and HTTP cache middleware. Gives concrete numbers: asking for fewer fields cuts 8 ms (4%) on a 200 ms call, but server-side caching can make endpoints 46% faster overall. Notes tools and standards that REST uses—JSON API, OData, JSON Schema, Varnish/nginx, HTTP/2—and warns GraphQL needs whole-data availability and struggles with caching and file uploads; ideal for read-heavy analytics.
AI-written summary · about
chapters (12)
- 0:00API bicycle analogy
- 2:00The hype around GraphQL
- 3:53REST vs GraphQL philosophy
- 5:25GraphQL features in REST
- 8:29History of query languages
- 10:26The importance of caching
- 12:57Database optimization
- 14:26Hypermedia and REST maturity
- 16:01REST as a state machine
- 19:00Implementing hypermedia
- 22:33Choosing the right tool
- 25:21Final recommendations
related talks
23mHarnessing the Power of GraphQL
30mWhy GraphQL vs. Why not GraphQL
54mGraphQL — Usable API or Just Another Hype?
18mIs GraphQL the end of RESTstyle APIs
29mFrom REST to GraphQL
39mWhy REST when you can GraphQL
33mGraphQL Will Do To REST What JSON Did To XML
51mGraphQL, gRPC or REST? Resolving the API Developer's Dilemma