RxJS — Reactive Programming Library for JavaScript

What RxJS Does

• Observable — lazy stream of values over time
• Operators — 100+ composable transforms (map, filter, merge, concat, switchMap, debounce, throttle)
• Subject — multicast observable, act as source and consumer
• Schedulers — control when subscriptions execute
• Error handling — catchError, retry, retryWhen
• Backpressure — sampling, throttling, buffering strategies
• TypeScript — full typing including operator inference

Architecture

Observables are lazy: they start producing values only when subscribed. Operators are pure functions that return new Observables. Pipeline compiles via the pipe() helper. Schedulers let you move work to microtask/animation frame/etc.

Self-Hosting

Library only.

Key Features

• 100+ operators
• Lazy by default
• Cancellable subscriptions
• Multiple subject types (BehaviorSubject, ReplaySubject, AsyncSubject)
• Schedulers for timing control
• Error handling pipeline
• Tree-shakeable imports
• Full TypeScript types

Comparison

FAQ

Q: Aren't Promises enough? A: A Promise is a single value; RxJS is a multi-value stream (event sequences, polling, push notifications). Plus RxJS supports unsubscription (Promises don't).

Q: Steep learning curve? A: The hardest part is mastering the 100+ operators. Start with these six — map, filter, switchMap, mergeMap, debounceTime, catchError — which cover most day-to-day use.

Q: Why does Angular lean so heavily on RxJS? A: Angular's HttpClient, Router, Reactive Forms, and EventEmitter all return Observables. RxJS is a first-class citizen of the Angular ecosystem.

Sources & Credits

• Docs: https://rxjs.dev
• GitHub: https://github.com/ReactiveX/rxjs
• License: Apache 2.0