Task vs Resource & RxJS

Choosing the right reactive primitive is essential for clean Angular architecture.

1. Task vs Resource / rxResource

Angular 19+ introduces resource() and rxResource() for declarative data loading. Here is how ngx-task compares:

Feature / Intent	Resource (resource / rxResource)	Task (ngx-task)
Primary Intent	Declarative Data Fetching (Read)	Imperative Action / Command (Write/Execute)
Trigger Mechanism	Reactive signal dependency changes	User gesture (click, submit, explicit .run())
Typical Use Cases	Fetching user profile by userId(), loading page data	Form submission, Login, Payment checkout, File upload, Autosave
Concurrency Policies	Automatic restart on signal change	Explicit policies: drop, restart, enqueue, latest, parallel
Visual Timing Controls	None	Built-in pendingDelay & minimumPendingDuration anti-flicker
Progress Reporting	Not supported	Built-in progress tracking via context.reportProgress()
Retries & Attempt Counter	Manual reload	retryLast() and context.attempt counter

---

2. Task vs RxJS Flattening Operators

In traditional Angular applications, developers use RxJS Subject streams combined with flattening operators (exhaustMap, switchMap, concatMap, mergeMap) to control concurrency.

Here is how ngx-task policies map to RxJS operators:

ngx-task Policy	Equivalent RxJS Operator	Key Advantage of ngx-task
drop	exhaustMap	Exposes signals (pending, running), anti-flicker, automatic component cleanup
restart	switchMap	Automatic AbortSignal cancellation for Promises & Observables
enqueue	concatMap	Queue limit & overflow policy controls (maxQueueSize, overflowPolicy)
latest	Custom buffer(1) / switchMap	Skips intermediate queued items, keeping only newest
parallel	mergeMap(..., limit)	Concurrent execution limit signals & individual TaskExecution handles

Why use ngx-task instead of manual RxJS subjects?

1. No Manual Subject Boilerplate: No need to create Subject, pipe operators, subscribe, or manage subscriptions.
2. First-Class Signal API: Direct consumption in Angular templates without toSignal() or async pipe.
3. Comprehensive Lifecycle & State: Exposes running counts, queue counts, progress, attempt counters, and execution handles out of the box.