Brandon Dail

Raw speed is good, but not enough. A scheduler allows us to improve the perceived performance of our apps, despite its overhead. 

With Fiber, React has embraced a scheduling-first approach. This has enabled features like time slicing and Suspense–but at what cost? Is scheduling just a workaround for React being too slow? 

This talk will dive into the answers to those questions (and more), outlining the importance of scheduling not just to React but to the web in general.

2019

ReactEurope

Scheduling is the Future

React Fiber was a full re-write of React that will enable new and exciting patterns around control flow, which we've seen previewed with React Suspense. But what is a fiber? How does it relate to a coroutine? What are algebraic effects, and why do I keep hearing about them? This talk will go over these computer science topics in the context of React Fiber, to help shed some light on how React Fiber is implemented and the control flow concepts behind the new APIs.

2018

React Rally

Algebraic effects, Fibers, Coroutines. Oh my!

Let's face it, the code we write very rarely resembles the code we send over the wire. Modern web applications are already built with ES6/ES7+, type annotations, JSX, and other language extensions. We treat JavaScript more like a target language: something we compile to, not something we write. If we’re going to treat JavaScript as a compilation target, we should start thinking critically about what we really want in a source language. In this talk, I’m going to try to convince you, through critical ReasonML’ing, that OCaml + ReasonML is one of the most attractive, up-and-coming options to fill this role.