Table of contents

A reactive value container that can change over time and track diffs between sequential values.

Signals are the foundation of the @tldraw/state reactive system. They automatically manage dependencies and trigger updates when their values change. Any computed signal or effect that reads from this signal will be automatically recomputed when the signal's value changes.

There are two types of signal:

  • Atomic signals - Created using atom(). These are mutable containers that can be directly updated using set() or update() methods.
  • Computed signals - Created using computed(). These derive their values from other signals and are automatically recomputed when dependencies change.
interface Signal<Value, Diff = unknown> {}

Example

import { atom, computed } from '@tldraw/state'

// Create an atomic signal
const count = atom('count', 0)

// Create a computed signal that derives from the atom
const doubled = computed('doubled', () => count.get() * 2)

console.log(doubled.get()) // 0
count.set(5)
console.log(doubled.get()) // 10

Properties

lastChangedEpoch

The global epoch number when this signal's value last changed.

Note that this represents when the value actually changed, not when it was last computed. A computed signal may recalculate and produce the same value without changing its epoch. This is used internally for dependency tracking and history management.

lastChangedEpoch: number

name

A human-readable identifier for this signal, used primarily for debugging and performance profiling.

The name is displayed in debug output from whyAmIRunning and other diagnostic tools. It does not need to be globally unique within your application.

name: string

Methods


get

Gets the current value of the signal and establishes a dependency relationship.

When called from within a computed signal or effect, this signal will be automatically tracked as a dependency. If this signal's value changes, any dependent computations or effects will be marked for re-execution.


getDiffSince

Gets the sequence of diffs that occurred between a specific epoch and the current state.

This method enables incremental synchronization by providing a list of changes that have occurred since a specific point in time. If the requested epoch is too far in the past or the signal doesn't have enough history, it returns the unique symbol RESET_VALUE to indicate that a full state reset is required.

Parameters

NameDescription

epoch

number

The epoch timestamp to get diffs since

Returns

Diff[] | RESET_VALUE

An array of diff objects representing changes since the epoch, or the unique symbol RESET_VALUE if insufficient history is available


Prev
Reactor
Next
BaseRecord