Discussing Backend For Front-end
summary
Argues that device-specific aggregation belongs to a Backend‑For‑Frontend team to avoid exploding client calls and duplicated per-service device logic. Demonstrates a Flask prototype that aggregates products, news and debug info, then moves that logic into an API Gateway plugin using API6 on nginx/OpenResty with Lua, deployed via Docker Compose and etcd or static config. Curl returns identical filtered JSON, proving the pattern works as an organizational—not purely technical—solution.
AI-written summary · about
chapters (17)
- 0:00Introduction
- 3:55Frontend-backend evolution
- 6:23Generic data delivery issues
- 7:06Microservices hype critique
- 10:01Conway's law reality
- 12:29Org barriers to microservices
- 17:01Microservices downsides
- 18:10BFF pattern explained
- 21:04BFF demo with API Gateway
- 23:59Apache API6 intro
- 25:05BFF with OpenResty
- 28:56BFF alternatives
- 29:18Against microservices
- 30:23Org problems beyond tech
- 33:37Counterproductive rules critique
- 34:00Closing remarks
- 35:01Q&A
related talks
37mDiscussing Backend-For-Frontend
26mBackend For Frontend: Building APIs w. end user in mind
21mEnhance and improve user experience with BFF
17mBackend for Frontend – The secret of a great mobile project
34mArchitektury Aplikacji: The last mile of microservices
44mMicro Frontend: the microservice puzzle extended to frontend
09mKeep your friends close and your APIs closer
39mScaling Frontend Development