There Is No Server
summary
A Rails blog is transpiled Rails blog and runs entirely in the browser from static GitHub Pages files: ERB, controllers and Active Record become JavaScript, breakpoints hit in DevTools, Action Cable uses BroadcastChannel, and an in-browser editor with hot reload works. Deployments give every user gets their own SQLite database, system tests run in ~75ms without a full browser, and throughput climbs from ~3,000 to 3,000 to 10,000 requests per second with alternative backends while memory drops from ~369MB to ~12MB; known limits include method_missing and C extensions fail.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 1:39Rails on a phone
- 6:58Ruby for DSLs
- 9:37Live demo: serverless blog
- 12:06Performance benefits
- 14:14Memory savings
- 16:44Collaboration with Matt
- 18:49Rails speed advantages
- 20:33Whole program inferencing
- 21:57Future plans
- 25:13Ruby as JavaScript
- 26:17Conclusion
related talks
26mFrom Server to Client: Ruby on Rails on WebAssembly
25mThe future of Rails begins in the browser
36mRails is Obsolete (But So's Everything Else)
28mBuilding in Rails, Backbone, and CoffeeScript
33mServices and Rails: The Sh*t They Don't Tell You
31mFrom Rails to the Web Server to the Browser
29mRails to Phoenix
33mYou'll Never Believe Which Web Framework Powers Upworthy