MULSID

Mini Unique Lexicographically Sortable ID. A short (10 character) ID for improved database efficiency in inserts and storage. MULSID is not suitable for certain use-cases, as it makes a compromise on randomness to favor shorter ID length.

About

MULSID uses a bit-packed layout internally. The 10-character base62 string encodes a single integer containing:

• Timestamp tick (upper bits) — derived from the current time using 9 ms increments
• 18-bit randomness (lower bits) — 262,144 random states per tick

This gives approximately 913 years of unique ID coverage from the Unix epoch.

You can read more about MULSID in this blog post. Check out the demo to see MULSID in action.

Install

NPM

npm install @mpact/mulsid

Bun

bun add @mpact/mulsid

Deno

deno add jsr:@mpact/mulsid

Usage

To generate a random MULSID, you can import and use the mulsid function

import { mulsid } from "@mpact/mulsid";

const id = mulsid() // 3pPgiiTJ9J

You can also guarantee a monotonically increasing ID by using the monotonicMULSID function

import { monotonicMULSID } from "@mpact/mulsid";

const id1 = monotonicMULSID(); // 3pPgn9dUpT
const id2 = monotonicMULSID(); // 3pPgn9dUpU (Always lexicographically greater than id1)

You can also determine the time the MULSID was generated within 9 milliseconds using the decodeTimestamp function

import { mulsid, decodeTimestamp } from "@mpact/mulsid";

const id = mulsid(0); // 0000000vaS
const time = decodeTimestamp(id); // 0

const id2 = mulsid(4); // 0000000tQq
const time2 = decodeTimestamp(id2); // 0 (the tick rounded down to the nearest 9 ms boundary)

const id3 = mulsid(10); // 00000018nH
const time3 = decodeTimestamp(id3); // 9