Project Summary

• Date: Feb 26, 2022
• Description: A client-side browser fingerprinting library that generates a unique device identifier by collecting 18 browser/system signals (all in a single vanilla JS file with zero dependencies).

Motivation

The goal was to prevent spammers from creating multiple accounts. I didn't like the results of the popular JS package FingerprintJs (free version) so I built my own Fingerprint JS library.

Links

Github repo: https://github.com/benlhachemi/fingerprint-id-jsFile.git

Tech Stack

• Language: JavaScript (ES5 with minor ES6 additions)
• Runtime: Browser (client-side)
• Dependencies: None (zero external libs)
• Module System: UMD (CommonJS + AMD + global script tag)

Key Features

• Canvas fingerprinting (renders text to canvas, exports toDataURL)
• Installed font detection via width comparison (~200+ fonts checked)
• User-agent parsing for OS, browser name, and engine detection
• Plugin/ActiveX enumeration (IE and modern browsers)
• Screen resolution, color depth, timezone, language, touch support, connection type
• Cookie, Java, Flash, Silverlight availability checks
• Legacy browser support (IE4+, Netscape 4, Opera 7+, etc.)
• Defensive try/catch pattern on every signal collector

What I Learned

• Classic browser fingerprinting techniques (canvas, font list enumeration, UA parsing)
• UMD module pattern for cross-environment compatibility
• Legacy browser quirks and fallback handling (IE ActiveX, historic Navigator APIs)
• How to implement a font detection algorithm using rendered element width comparison
• Building a library with zero external dependencies