The two ways to manage state
summary
Argues there are two workable approaches—direct mutation for speed and indirect event-driven/statechart actors for maintainability—and shows concrete tradeoffs: mutable state is faster, proxies/signals let code pretend to mutate, effects belong with state, and single-store lies. Uses an espresso machine demo wired to an actor/statechart model and calls OpenAI to compute shortest paths to make drinks; the AI route works about 90% of the time after a shaky Wi‑Fi run, and the library shown is xdate v5 beta.
AI-written summary · about
chapters (6)
- 0:00Intro
- 1:00Managing State
- 5:46Immutable State
- 10:38MultiStore vs SingleStore
- 16:35Everything is a graph
- 18:55Demo
related talks
33mThe Actor Model, Behind the Scenes
29mReactive State Machines and Statecharts
24mPanel discussion about state management
21mReactive State Machines and Statecharts
01h 01mThe visual future of reactive applications with statecharts
52mState management beyond the libraries
37mReactive State Machines and Statecharts
25mMastering UIs with Finite State Machines