15 Years of Forking

Fifteen years ago today, I posted a thread on the Overclock.net forums. I was sixteen, I had an HP Compaq TC4400 that I’d convinced my parents would “improve my school work”, and I was frustrated that Firefox didn’t have an official 64-bit build. So I compiled one myself, called it Waterfox, stuck it on SourceForge and went back to my A levels.

Within a week it had 50,000 downloads, completely unexpected. Frustratingly, being on an island in the Mediterranean meant there was no support network or anyone to turn to with regards to “what’s next”. Had I been stateside, with the infrastructure and institutional knowledge of “tech”, who knows - I might’ve had a guiding hand on how to manage something like this and work with the momentum. But alas, I would have to learn a lot of painful lessons myself.

Fast forward to today, 15 years later, and Waterfox is still here. So am I, albeit a bit older and significantly more tired. At best estimates, Waterfox probably has around 1M monthly active users.

Where it started

If you go and look at that original OCN thread, it’s a very different world. People are talking about Silverlight support, MSVCR100.dll errors, and Peacekeeper benchmark scores. Someone asks for a 64-bit Chromium build and the thread title gets updated with every new Firefox version, all the way up to 56.0.2.

Originally, and under the username MrAlex, I was only trying to earn enough forum reputation so I could trade and buy second hand PC parts. I didn’t have a plan and I certainly didn’t have a business model. I just thought it was cool that you could take someone else’s source code, compile it with some changes, and end up with something different. Open source is a wonderful thing when you’re sixteen and don’t know anything about the software development lifecycle, yearning for knowledge.

What happened in between

You can scour the internet, read this blog or view the media carousel at the bottom for then until now, but the short version: Waterfox grew - a lot - over 25 million lifetime downloads, and that figure is from calculations about seven years ago so the real number is certainly higher. I went to university, studying Electronics Engineering at York before a masters in Software Engineering at Oxford. I tried to start a charitable search engine, which failed as badly run startups tend to do. Ecosia reached out and something nice happened - Waterfox users helped plant over 350,000 trees in a single year.

Then System1 came along. I joined them, served as VP of Engineering, and helped scale the browser engineering team through a NYSE IPO - a genuine education, though companies change and focus shifts.

So I took Waterfox back under BrowserWorks, independent once again. The three years since have been simultaneously the most difficult and the most rewarding of Waterfox’s existence.

The hard bit

I’m not going to pretend the economics of running a privacy focused independent browser are great, because they’re really not. When Bing terminated all third party search contracts it hit hard - search partnerships are basically how independent browsers survive, and revenue has been poor since. There have been a few months in the red recently.

Other ways browsers make money just feel icky, and it’s not something that Waterfox stands for either.

But, pain and all, I keep coming back. Every time I think about stepping away, someone sends a kind message through the donation page, or I see a thread somewhere of someone discovering Waterfox for the first time and being pleasantly surprised. There’s a community here that cares, and I care about it.

I want users to know that whatever future steps I’ll take, they’ll always be with Waterfox and its sustainability in mind.

What Waterfox is in 2026

This year will see Waterfox shipping a native content blocker built on Brave’s adblock library - and it’s worth explaining what that means and why.

The blocker runs in the main browser process rather than as a web extension, which means it isn’t subject to the limitations that extension based blockers like uBlock Origin face. It’s faster, more tightly integrated, and doesn’t depend on a separate extension process or require us to constantly pull in upstream updates. Brave’s adblock library is also mature - it has paid engineers working on it, a wide filterset, and crucially it’s licensed under MPL2, the same licence as Waterfox, which makes it a natural fit. uBlock Origin, as good as it is, carries a GPLv3 licence that would’ve created real compatibility headaches.

For how it works in practice: by default, text ads will remain visible on our default search partner’s page - currently Startpage. The idea is that this is what will keep the lights on. Users who want to disable that entirely can do so with a single toggle in settings, and it has nothing to do with any of Brave’s crypto or rewards ecosystem - we’re just using the adblocking library. Everyone else gets a fast, native adblocker out of the box, no extension required.

If you already use an adblocker, don’t worry, you can carry on using it. This will be enabled for new users or users who aren’t already using an adblocker.

In the meanwhile, Waterfox’s membership of the Browser Choice Alliance alongside Google and Opera, is pushing for fair competition and actual user choice in the browser market.

And we still don’t have AI in the browser. That hasn’t changed. The browser’s job is to load web pages, keep your data private, and get out of the way. It seems other browsers have forgotten that.

Fifteen more?

I’d like to think so. The browser market is more diverse than it’s ever been in terms of soft forks - everyone and their mum seems to be launching a variation of Firefox. Running an independent browser is getting harder, not easier. But there are more people who care about privacy now than there were when I was compiling a blue Firefox on a tablet PC in my bedroom. More people who want software that respects them.

Waterfox started because a sixteen year old wanted a faster browser. Fifteen years later, it’s still here because enough people want a browser that works for them - not for AI companies, and not for anyone else.

Thanks to everyone who’s been part of this - from the OCN community who gave those early builds a chance, to the people who send donations with messages that make my day, to the contributors who submit patches and file bugs. This project has always been bigger than me, even when I’m the only one working on it.

Here’s to the next 15!