Releasing an Open Source User Agent Lookup Table

We're building PageviewsOnline as a side project. The goal is simple: privacy-first web analytics that focus on pageviews, not people.

As part of building the product, we've released a user agent lookup table as an open source project.

It's used to turn raw user agent strings into basic, high-level data like browser and operating system. It's not meant to be perfect or overly advanced- it's designed to be fast, predictable, and good enough for simple analytics.

Why We Open Sourced It

• Trust and transparency - Analytics and user agents can feel sketchy. By open sourcing this, anyone can see exactly how it works. No hidden logic, no tracking tricks
• Performance over cleverness - The lookup uses normalized strings that can be hashed and compared directly. This is faster and more predictable than complex regular expressions, especially at scale
• JSON and language-agnostic - The dataset is stored as JSON, making it easy to inspect, version-control, and use from almost any programming language
• Simple by design - This is not a sophisticated detection engine aiming for perfect accuracy. It's built to be fast, reliable, and stable, while providing basic analytics that don't break as user agents change
• Community-driven - Because it's open source, anyone can use it, fork it, or contribute improvements. Our hope is that it improves over time through real-world usage and community input

The project is actively used in production, but it's also useful on its own. Open sourcing it felt like the right fit for how we want to build PageviewsOnline: simple, honest, and privacy-first.