Long-Time Slashdot Reader Announces New Open Source Web Browser 'Northstar'

Long-time Slashdot reader Andreas(R) announces Northstar, a "minimalist" web browser written entirely from scratch in C — now available in version 1.0.11: One window, one page, one process — about 160,000 lines of original C, small enough for one person to read and audit end-to-end. It carries no upstream browser engine: no Gecko, no WebKit, no Blink. It is free software under the GNU General Public License, version 3 or later, and it runs on Linux, macOS and Windows...

Compliance is not inherited from an upstream engine; it is measured, section by section, against the specification text... Private by construction. No telemetry, no update pings, and no AI assistant or AI-style web APIs. Safe browsing checks a top-level navigation's host against a local SHA-256 blocklist before it is fetched — the check never leaves your machine. Cookies, storage and the cache are partitioned by origin.
It's logo reminds me of the old Netscape logo. (And I love the "Best viewed in..." GIF at the upper-right of its web site.)

Uah no

[thegarbz]

What's the use case for this unless you work for the DoD? Even before we talk about horrendous javascript performance (though surprisingly good Acid scores), where it shows Google maps running at around a frame every 2 seconds when moving around the map, the UI doesn't scale so I literally can't read any text on a 4K monitor. I can't even really make out the symbols they used on the buttons.

Re:Uah no

[allo]

Everything starts small. You're probably comparing to 30 year old browsers (with a million open bugs in their bugtracker). If this is written from scratch without using code from other browsers, it is damn impressive what already works.

Re:Uah no

[Rendus]

Everything starts small. You're probably comparing to 30 year old browsers (with a million open bugs in their bugtracker). If this is written from scratch without using code from other browsers, it is damn impressive what already works.

"Claude, write me a web browser. Start with the core components required for a minimal viable product."

https://github.com/nordstjerne... [github.com]

From scratch, in a sense. Scratching deez nuts.

Re:

[gweihir]

So this thing is slop? Even more reasons not to use it. It means everything not specified in detail is suspect, missing, broken or insecure and maintenance will be an absolute nightmare.

Re:

[Anonymous Coward]

Who wouldn't use AI for starting a potentially huge new project? Don't let the boomers ruin your day, AI is now ubiquitous in SWE.

Re:

[dskoll]

I would not. But then again, I'm old-school and retired, so I get to develop the way I want, not the way some corporate bean-counter wants.

Re:

[Pieroxy]

I suggest you give coding agents a try. They really help. You need to keep them on a tight leash, but they help you move faster and often better if you know how to use them. Nothing related to corporate bean-counters.

Re:

[AmiMoJo]

I went to Subway and told the lady to give me a 6 inch chicken tikka on Hearty Italian (TM) with cheese and some sweet sauce. My wife was really impressed by the sandwich I made.

Re: Uah no

[backslashdot]

Yes, so eventually it will be just like Chrome?

Re: Uah no

[alexandre]

If you're going to slop and still picked C over Rust in 2026... Ouch.

Re:

[ChunderDownunder]

We already have a half-finished Rust browser - Servo!

Re:

[arglebargle_xiv]

For me it was looking at the list of updates, there's major new features being added at the rate of one every few days in some cases, there's no way a human is writing the code there.

Re:

[dfghjk]

So what? Your impression doesn't get any work done for me.

Re:

[thegarbz]

You're using this browser for work? Common, share your use case. I'm was asking you what this is good for, not asking you to critique the reason why it's not good for me.

But since you're here I doubt you're getting any work done anyway.

Re:

[gweihir]

The only thing impressive here is the sheer undiluted stupidity of the approach and of the ones cheering for it. Like you.

Re:

[thegarbz]

Everything starts small. You're probably comparing to 30 year old browsers (with a million open bugs in their bugtracker). If this is written from scratch without using code from other browsers, it is damn impressive what already works.

It's not a question of starting, my question was "what is driving the need?"

While it's good to see a browser work, it's a shame that the code base seems to be focused on something designed to be as portable as possible rather than functional. Any idiot can cobble together 2 lines of code which actually do support proper UI scaling, but right now they've released something non-functional. I'm not questioning bugs here, I'm questioning their approach.

Re:

[allo]

1) Take the code
2) cobble together the 2 lines,
3) ???
4) PROFIT

Re:Uah no

[Andreas(R)]

The case for this browser, is that there needs to be multiple independents web browsers, to push for innovation. Google Chrome is NOT open source, and is bloated and monopolistic. Competition drives innovation!

Northstar browser is implements in C, takes a few minutes to compile and is secure because it doesn't rely on a JIT compiler.

Re:

[dfghjk]

"...and is secure because it doesn't rely on a JIT compiler."

LOL SuperKendall is that you?

Re: Uah no

[fortfive]

Also, pushes standards compliance. And (pipe dreaming) with wide adoption might result in downscaling of all the layers of web 2.0 malarky that slow everything down so much.

Re:

[ambrandt12]

Andreas(R)... but it works.
It's funny how everyone talks about 'secure' browsers... the browser can be Fort Knox, but if you click on the wrong link or download the wrong thing, you'll learn how secure it is.
Not to mention, how does it handle stuff like YouTube or Discord or Zoom? Does it handle plug-ins like AdBlocker?

Re:

[gweihir]

And if the browser is NOT secure, you have several times the problems. Your point?

Re:

[ambrandt12]

What is your metric for secureness?
You do know that none of the governments of the world are gonna allow some encryption system that renders your IP-related encrypted communications without those governments having the master key.

Re:

[AmiMoJo]

A better use of those AI tokens would have been to re-work Firefox into something that is easier to build and work on. The code is horrible and difficult to even compile, which really holds Firefox back and keeps everyone using Chromium as the base of their browsers (Edge, Brave, Opera, Vivaldi etc).

Because nobody is going to use this browser daily and the only way to reduce the dominance of Chrome is to get more browsers to switch away to basing themselves on Firefox. At the moment every Firefox "fork" is

Re:

[gweihir]

Forget it. This person threw together what they mistake for a good spec of a browser and then asked Claude to build it. They do not have the skills to do any real work, LLM support or not. This whole thing is just yet another meaningless stunt that impresses the clueless.

Re:

[thegarbz]

So DoD like I said? Precisely the security over usability justification you just mentioned. The issue is if you don't end up with a user base a project often becomes stale and defunct. I hope this doesn't happen.

Also we have multiple rendering engines already. I don't see one not supporting basic functionality such as JIT compiling (which is largely credited for actually making the internet usable with the modern functions we expect) as a feature; users will consider this a bug leading to poor performance a

Re:

[Zero__Kelvin]

Chrome isn't Open Source, but Chromium [wikipedia.org] is the Open Source code that Chrome is based upon.

Re:

[dgatwood]

...and is secure because it doesn't rely on a JIT compiler.

I believe you misspelled "slower than molasses in February".

Re:

[oddtodd]

+2 Funny

Re:

[votsalo]

Northstar browser is implement[ed] in C

Would you consider adding the Claude prompts in the repository, or in a separate repository? They might be more interesting than the C code.

Re:

[gweihir]

Come to think of, the whole article sounds very much like slop.

Of course the claim that secure or not is a question of using a JIT compiler is completely bogus and just demonstrates that the one making it has absolutely no clue about software security.

Re:

[drinkypoo]

Classic Slopdot, I got modded down for my comment while he's using Claude to "write" the code.

Re:

[taustin]

The exclamation point at the end is a clue.

Re: Uah no

[guacamole]

It's the NCSA Mosaic of 2026.

Very cool

[AcidFnTonic]

This is very cool, will be playing with this today.

Wow, thank you Slashdot!

[Andreas(R)]

Available for Windows on Windows store [microsoft.com]
Git repo https://github.com/nordstjernen-web/nordstjernen-browser [github.com].
There are 3 variants of the browser now, so we can do experiments and research. We also need funding. All variants are available open source.

We are 3 developers now, a Polish man, a Spaniard, and me in Norway.

Re:

[robot5x]

hi Andreas - well done for doing something, instead of just sniping from the sidelines and shitting all over someone's efforts like vast majority of idiots in this thread. As a non-member of the Computing Purity League I wish you all the best.

Re:

[SlashbotAgent]

If it were Claude, it would be 160,000,000 lines of code. And broken.

Re:

[aitan]

Look at the commits
https://github.com/nordstjerne... [github.com]

Re:

[SlashbotAgent]

Doh!

Dillo

[markdavis]

According to the site it has no support for:

* Tabs
* Multiple windows
* Modern video
* WebGL
* WebGPU
* Extensions/Plugins

It reminds me of a more modern Dillo https://en.wikipedia.org/wiki/... [wikipedia.org] except with Javascript and more CSS. It could be useful for appliances that just need an HTML viewer, access to a Wiki, documentation viewer, or for testing.

It is a small subset of https://nordstjernen.org/ [nordstjernen.org] which is much more capable (but much larger/more complex).

That means no ad blocking

[RitchCraft]

A browser that does not support ad blocking is a non-starter. You simply can't use today's Internet without ad blocking in place.

Re:

[markdavis]

>"A browser that does not support ad blocking is a non-starter. You simply can't use today's Internet without ad blocking in place."

I agree. It is why you should use Firefox + UBO, or one of its forks (LibreWolf, WaterFox, whatever). But that is why I said the use case is for things like a wiki, documentation, appliances, or for testing. The same type of things one might do with Dillo.

Why not

[Tailhook]

AGENTS.md CLAUD.md .agent/skills .claude

Commits on Oct 3, 2026
claude committed yesterday
claude committed yesterday
claude committed yesterday
claude committed yesterday
claude committed yesterday

I suppose as long as we have these mass license+copyright bypass onions at our disposal, we may as well use them to make alternative browser engines?

Sure. Why not.

Re:

[truedfx]

It's one thing to do this in the first place, it's quite another to call this "written from scratch".

Re:

[Ksevio]

What would you call it then? Doesn't look like it's built off an existing code base

Re:

[Ksevio]

I guess all code is based on existing code bases if that's how you want to go with it. Nothing is written from scratch except people that have never been exposed to code before.

Not the view most people would go with but it has a certainty to it for sure

Re: Why not

[robot5x]

Eh? Wtf are you talking about?? It commits to git and you release under a licence. I can fork that repo and quite easily pass it to codex or Gemini or whatever, or even build on it myself. I get there are a lot of AI haters in here, but try and pick an angle that actually makes sense.

Sorry, but no

[gweihir]

The overwhelming concern for any browser these days is security. And that means looking at all the libraries that get pulled in as well. And suddenly things get a lot more murky. Until we get real defense-in-depth as a mandatory requirement in all software that can be taken seriously, this is an arms-race and a single person cannot compete.

Re: Sorry, but no

[Andreas(R)]

We are 3 developers now for the main browser.

Re: Sorry, but no

[gweihir]

No chance. And since you start with slop, completely forget about it. Nice hobbyist demo project, but that is it.

Re:

[ArchieBunker]

You haven't seen what Claude is capable of.

Re:

[ambrandt12]

Sure, it'll carry out your prompt ("Clod, make me a minimalist browser"), now you don't even have to know how to code!
Soon, your fridge will be sending you texts and emails to tell you that your milk will expire in 43 1/2 hours.

Re:

[gweihir]

You have no clue what you are talking about. Like so many incompetents that cheer for LLMs.

Re:

[Ksevio]

The people calling all AI output slop are the ones that messed around with LLMs a couple years ago and then didn't keep up with them. Most new software now contains code generated by LLMs

Re:

[gweihir]

You are arguing on the level of "you are holding it wrong". Pathetic and marks you as not very smart. Which, incidentally, makes you the primary LLM target group.

Re:

[SCVonSteroids]

Look, it's blatantly obvious you hate LLMs, we know, and that's completely fine, but you can't re-word someone's rebuttal to you into total nonsense and then proceed by calling them "Pathetic" and "not very smart". If anything it just makes you sound like one orange man we all know too well (unfortunately). But you do you.

Re:

[Ksevio]

There are lots of tools I don't know how to use which are complicated and require some learning to use. For example I've never learned the 3d modelling software Blender and opening it looks hard to use. Does that make people who know how to use Blender and find it helpful that tell me to read the manual "pathetic and not very smart"?

Re:

[allo]

I'd wish for the opposite: Suppose a modern browser would use libraries for everything. A video element? Use the appropriate system libraries. Canvas? Could probably be implemented using some Gtk/Qt drawing primitives. And then you don't need to update a 200 MB browser each time a single component needs an update, but your usual system-wide updates also update the relevant browser components, without changing the overall browser experience. Browsers could also share these libraries. Why does each browser ne

Even the description is AI written.

[Rendus]

The entire thing has the funk of bullshot. Yay.

Re:

[NobleNobbler]

Came here to say this... been using Claude enough to smell it's descriptive style a mile away...

"Compliance is not inherited from an upstream engine; it is measured, section by section, against the specification text... Private by construction. No telemetry, no update pings, and no AI assistant or AI-style web APIs. Safe browsing checks a top-level navigation's host against a local SHA-256 blocklist before it is fetched — the check never leaves your machine"

Can't overstress how much I hate this

Re:

[timeOday]

Any reason to disbelieve it or the benefits? Or just emotionally-centered anti-AI-sentiment?

Re:

[DigitalSorceress]

this... tracks.

I've been using Claude Code at work for a low-stakes project so that my company sees me being all "AI forward" and stuff (They're really drinking the Kool-aid) and yeah I'll admit its been doing OK on the project.... and yeah I see where you're coming from, there is a certain prose about how it outlines/summarizes things that feels familiar here.

I would really like a true new browser that doesn't use WebKit, Chromium or Gecko (and don't get me started on Goana) but I question such a tiny team

I genuinely wish you the best of luck!

[cascadingstylesheet]

Modern JavaScript and CSS are insanely complex ... and are not just for silly decoration, but for needed functionality and things like WCAG compliance.You sure have your work cut out for you.

I very much want real alternatives though, and I do genuinely wish you success!

Maybe a DamnSmall Linux browser candidate.

[ClassicASP]

More than once I thought to myself: "Ya know, way back in the day, it was wise to write source code such that it was as small and lightweight and efficient as possible in terms of excecution effort and RAM usage and driver read-write activity because of the hardware limitations we had to deal with. As hardware evolved we became less and less concerned with that. Wonder what would happen if we just started over and took today's modern hardware and took the old-school super-lightweight approach to software

Re:

[ClassicASP]

.....to clarify: metaphorically sort of like owning a high end sports car and putting Shell V-Power Nitro Premium grade fuel in it instead of regular unleaded.

Re:

[dfghjk]

Or having a multi-core system and running the world's most tightly coded single-threaded application on it. Performance is more than good wishes.

Re:

[lucifuge31337]

Maybe the code you strat with shouldn't be AI slop like this project.

Re:

[dfghjk]

"As hardware evolved we became less and less concerned with that."
That's not true, your problem is who you think "we" is. For application programmers, arguably there hasn't been concern since the early 90s. But the majority of programming is embedded where constrained systems make you care about resource usage.

"Do we get lightning speed performance?"
No, lightning performance results from care and good work, not hypotheticals.

"There's no point in doing that without a minimalist browser that can still surf

Re: Maybe a DamnSmall Linux browser candidate.

[Tomahawk]

Yeah, it is a little grating that software tends to feel like it's running "at the same speed" as software from the 80s or 90s. As hardware gets faster and faster, software seems to just get more bloated to compensate. Frameworks and libraries can be too blame for a lot if it. But also the lack of need to optimise anything. I'd love to see a modern (young-ish) developer trying to write something for the 32k machines of the 80 (C=64, say), or even the 512k machines like the Amiga and Atari.

That would b

Go a step further to better free&open standard

[Paul Fernhout]

I know there is an XKCD about standards (and adding yet one more to create yet more confusion).
"Standards"
https://xkcd.com/927/ [xkcd.com]

But that said, HTML, CSS, and JavaScript are all terrible standards in many ways because of their complexity (and major design flaws like JavaScript having vastly too many global variables). Asa an alternative, what if, say, we all just used something like Scheme (or Common Lisp) to encode web documents and web code as S-expressions in some semantically standardized formats? Scheme

ehh

[nomadic]

I mean the software is impressive and all, but since when is someone with a user ID in the 400,000s a long-time reader?

Re:ehh

[Baron_Yam]

It happened when you got old and apparently didn't notice.

BTDT.

Re:

[dfghjk]

That doesn't consider how many times he's changed his ID due to personal humiliations.

Re:

[strikethree]

I mean the software is impressive and all, but since when is someone with a user ID in the 400,000s a long-time reader?

This account has been active since 2004 or so. That counts as a long time. 20+ years definitely qualifies as a "long time".

This is amazing

[WaffleMonster]

Just getting the stack up to render all of this shit is cool as shit. Maybe try and look for niches browser vendors are too corporate to touch and do them anyway?

Something I think could eventually have a huge impact if it ever caught on is diversifying away from PKI and the curse of the ubiquitous adhoc web login form... Support for whatever PAKEs are in OpenSSL like TLS-SRP. These let you login to websites securely without any certs via secure phishing resistant mutual password authentication bound to the TLS layer. If I remember right at least Apache and cURL support it but no browser does and needless to say no publicly available website I know of does either.

Other details

[NotEmmanuelGoldstein]

North Star binaries available from Microsoft Store:
https://apps.microsoft.com/det... [microsoft.com]

Based on the minimalist browser for linux:
https://github.com/nordstjerne... [github.com]

The author has started a third version based on a different JavaScript engine, which is incomplete:
https://github.com/nordstjerne... [github.com]

Why not Rust?

[Tomahawk]

You're writing this new in 2026. You're using AI Agents to do a lot of the work, it seems, based on the files in the repo.

And you are writing it in C ??!!?

Now, I love C -- I love that you can do anything at the hardware level, and do crazy stuff with pointers, and all the low level goodness. It's a fun language, seriously powerful, fast.

But C makes it notorious easy to introduce a plethora of security issues! Not matter how hard you (or your agent) try, they WILL get in there. Buffer overruns being the simplest to exploit, and the simplest to accidentally include (and can be hard to find too).

So why not Rust? Why not just take the easier option to safety. Especially with something that you want many other people to be using. Rust is powerful, secure, fast. I really can't see a reason to not use it for a project like this.

I daresay it's not too late to have your agent rewrite everything in Rust. It's 160kloc. It won't take it long.

Re: Why not Rust?

[Andreas(R)]

I understand C well, I don't know Rust. C is like the latin of programming languages. That's why I chose C.

Re:

[Nahor]

Latin in a dead language, so, maybe, it's time to learn something a bit more modern? :p

Re:

[thegarbz]

Do you prefer reading Caesar, Cicero and Ovid in English ? I trust not ...

I think very few people except for true language experts have the ability of stating any kind of preference on this matter, which is sort of the point.

Re:

[thegarbz]

C is like the latin of programming languages.

I mean no disrespect to your efforts, but you are right. Latin is a language used for teaching of basic structure of it's modern successors. People learn it and yet no one actually uses it for anything new. C like Latin should really remain an academic curiosity in the modern world unless you're working on a system anchored with a long history and have the brightest language experts dealing with it (like the Linux kernel).

Re:

[manu0601]

C like Latin should really remain an academic curiosity in the modern world

You know there is still a lot of C running everywhere, right?

Re: Why not Rust?

[Tomahawk]

But if you are using an AI agent, does the language matter? Plus it's an excellent way to learn Rust.

Re:

[rickb928]

And a new meme is born. 'Imagine this written in Rust'...

On a Beowulf cluster.

Re:

[manu0601]

And you are writing it in C ??!!?

We all know the dangers of C, but I still prefer a program written in C by someone that masters that language, rather than in Rust by someone that does not know what is going on.

Re:

[ChunderDownunder]

A FAQ of sorts [nordstjernen.org] explains their reasoning behind C - using the latest C23 spec to avoid many of the historic pitfalls and having Claude scan the code for any incorrectness.

So yeah C is well understood if you trust Claude to proactively prevent any nasties, perhaps.

Re: Why not Rust?

[robot5x]

Because then everyone would be bitching about him using AI *and* Rust.

North Star...

[NobleNobbler]

I've been seeing this term pop up a few times lately-- are the LLMs starting this trend?

How does Claude and GPL make sense together?

[fleeped]

Can anybody explain? So we can use the plagiarizing genAI backend to generate software, then we label it as GPL, which is then ignored by other plagiarizing genAI crawlers, and is used to limit real developers? Why use GPL? Is it important that no other besides AIs can benefit commercially from the code? Might as well be public domain, as all AI-generated code should be.

Re:

[fleeped]

Ok incorrect assumption here, that GPL hurts profitability, bit of a brainfart really now that I think of it

Re:

[drinkypoo]

Can anybody explain? So we can use the plagiarizing genAI backend to generate software, then we label it as GPL,

Can a person apply a license to code they don't have the copyright to? The answer is no. This code is either infected with a mishmash of copyright from its original sources, or it's not copyrightable at all, and courts haven't decided which but they are both bad.

Re:

[fleeped]

What you say makes perfect sense, however it seems that with genAI-washing (or code laundering?) we're entering, tumbling, in a post-licensing era.

Re:

[drinkypoo]

To my mind, nobody wants this to get to this supreme court, because the outcome could be anything someone is willing to pay for. But it's going to wind up there eventually. If it does get there under this administration the most likely outcome is something beneficial to AI service providers, quite possibly allowing GenAI to wash away copyrights. But will it do so enough to allow someone else to copyright the result, or will it be basically uncopyrightable and therefore (in the case of code or other, similar

AI slop, now as a browser?

[moglito]

The entire description sounds Claudish, and "160,000 lines of original C, small enough for one person" is utter BS. Must have been written by AI and said human who is claiming to read all that code has probably never worked on a project nearly as big. 160k LOC per person is completely unrealistic. Already at 20k most devs become very slow, i.e., can no longer remember how all of it worked and need to refamiliarise themselves with parts of the code when they review changes to it.

The code is self-explaining.

[bobm]

From the repo: The code is self-explaining. Don't write code comments.

since you aren't writing the code why not ask the AI to add in comments?

Minimally I like a header for each function but that's because around the time Java hit the ground we seem to have lost the art of the man page.

Author's second go at vibe-coded slop browser.

[BenFenner]

This browser is vibe-coded slop barely capable of much of anything, and is the author's second try in about as many months.
They have been spamming us over at the Pale Moon forum for months and have barely avoided a total ban.
There is nothing to see here.

Vibe-coded browser attempt #1 [palemoon.org]

Vibe-coded browser attempt #2 [palemoon.org]

Tiny menu, doesn't render anything I read

[nospam007]

Useless, deleted.

LadyBird

[paugq]

Why should I use this NorthStar browser instead of LadyBird?

Re: North Star hunh?

[Ploulack]

I think it's mou (long).
I thought that was a more appreciative comment than a smiley.
Unsure. :)
Go Ken!