I wonder what Evan Wallace would think

Luke Smith enters the chat

Yes! This! I'm in!

You can just fork something. Starting from scratch is foolish unless you're genuinely just trying to learn things

You'd be better off using and contributing to Firefox.

Less work, more result.

The spelling mistake in the first sentence really throws me off. I'm not trying to be overly picky or pedantic but after reading only 7 words into a multiple-thousand word long post and seeing an error, I begin to think that this may be a tough read. I also subconsciously begin to assume that the browser this person wrote must be riddled with bugs. Whether that's fair or not.

"This blogpost starts with me switching of my car radio..."

of != off

> And how about browser-specific extensions that make sites unusable from other browsers (sorry, you cannot view your gmail with Firefox. Please install google chrome)? This has happened, and as soon as lawyers find a way out of monopoly issues, it will happen again.

If anything would rile the regulators, Google making Gmail compatible only with Chrome would do it.

I think someone can make their own browser not to take over the browser market or compete with the giants.. it would be cool if we had tools to all make our own bespoke browsers tailored to our individual tastes. I don't think it would change the development side of things too much as the the big four would continue to be the target platforms.

OP, love it. This is the spirit that makes change.

I’m sure it will batter your brain and crush your soul as these large projects tend to do, but declaring war on a problem and then getting featured on HN doing so is old school hacker and you deserve street cred for it.

Congratulations on your new baby. Can’t wait to see what college you choose! ;)

Great! Here's a timeline of web browsers, both open source and closed, btw:

https://upload.wikimedia.org/wikipedia/commons/7/74/Timeline...

Good for you!

I've been considering doing something similar. I can't find a browser that really makes me comfortable or happy, so my thought is that the only solution is to write my own.

It would certainly not be full-featured or be able to deal with the full HTML spec -- but for me, that's a positive, not a negative.

I fully support it.

Qs: wouldn’t an, potentially better, approach be to use an existing rendering engine, say WebKit or Gecko, and then build the rest on top first. And, then decide if the needs are met and then decide to rewrite or fork the rendering engine?

Disclaimer, I don’t have a deep understanding of browsers.

10 years from now: "GPT-12, write me a browser. Make it like the latest chrome, but make it not do this annoying thing. Build me a github site with full documentation. It should work on Mac, Linux and Windows and support all the latest standards."

I'd rather advice people to reinvent a new html js format.

Preferably serialized to binary so it can fast and save battery on mobile.

Html was not designed for applications, it carries a lot of debt, and it's an ambiguous language, which makes it very expensive to parse for browsers.

A thought, write a very constrained sane rendering/layout engine as a compile target for an llm to convert html/css to rather than support all css or any of it for that matter. Could there be some kind of llvm for html/css?

Ambitious project, I'm sure you'll learn tons with it. Best of luck. :)

I hope you stick with it.

I would love a browser that sticks to the original use, to render simple html documents.

No Javascript or bloated features. Would break a lot of websites but that is ok.

It might even be good for accessibility on some sites

As someone who is annoyed with the direction the web is going, please keep at it. Not gonna lie, I also want to try building my own browser as a distant future project.

You can do it! I hope more people begin these kind of projects!

The only way this is feasible is if you find a way to code in a DSL and have AI generate all the implementation. A browser is extremely complicated.

Corporations are not hiring devs... Devs have plenty of time... Devs fuck corps by redoing things right... The shortsightedness of suits is as always staggering.

I'm glad to see this, and it looks like you're also building a new engine. I appreciate that and look forward to your success.

Exciting! Challenging for sure! Good luck!

Maybe consider hosting the source code in an open-source alternatieve to GitHub?...

"Now you have two problems."

Please integrate up to date XML technologies. In addition, please make it like the XUL based Mozilla.

i think this is a great idea, i hope you find others that are willing to pitch in

this isn't the first frustration-funded browser post i've seen, i think this is a common phenomenon. individual people want to gain control and this is a good sign.

So, obviously jaytaph is doing this for his own reasons, which is more than fair.

I've looked into the browser space, and its history, how we got here, and where else we might have gone, and might still get to.

Regarding browser history, there's the question of how we ended up with a single monolithic application which handles the many different elements of Web interaction, from HTML rendering to JS interpretation to identity and security to managing bookmarks and history, to multimedia (video and audio playback), and the often strongly conflicting roles of document reading and discourse versus commerce and transactions versus media playback versus general application support. (I've long thought that those four principle roles might best be split out.)

One of the more interesting (and obscure) early browsers was ViolaWWW, initially released in 1992, which was a graphical browser with scripting and stylesheet support, and introduced or provided an early exemplar of bookmarking, browser history, images, tables, frames, and other features that now define the modern browser.

<https://en.wikipedia.org/wiki/ViolaWWW>

<https://web.archive.org/web/20160303174535/http://www.viola....> (original seems dead: <http://viola.org/>)

I've long wondered why browsers don't offer the option to simply cache and store content locally rather than rely on remote lookups for all access. There are some sites for which this doesn't work especially well, but for many static sites, the ability to browse offline and search locally would be a powerful bonus. Similarly, why isn't it possible to run a search across content from my recent browser history rather than across the entire Web? If I know I'd read something in the past hour, or day, or week, constraining a search to recently-accessed content would be a strong bonus.

(How to deal with content changing over time by saving snapshots and diffs of websites could also be interesting.)

Or why isn't the notion of a discussion --- threaded, flat, sorted, weighted, whatever --- a first-class entity within the Web world? There have been an untold number of reinventions of what was essentially Usenet-plus-something, and the notion that discussions must be fully specified by each individual website (including HN) makes increasingly less sense to me.

I've toyed around with the notion of the Web as a filesystem-addressible space. Plan9OS with its 9P protocol has something like this. The notion is that Internet and Web space could be navigated as a filesystem using both existing and newly-developed shell and other tools to directly access and manipulate objects.

A modular approach to Web tools would mean that we could disaggregate HTML parsing, various proxy services, caching and storage, ad-blocking, privacy indirection (Tor, VPN, or other forwarding mechanisms), interactions (editing and such), among a suite of tools. Some of this now exists, but the hurdles to general use are still fairly high. Parking something like Readability.js directly in the mainstream content parsing flow would be an interesting option.

Working off Readability, I've had a notion that a set of predefined document types, to which user-selected styles could be assigned, would be an interesting change-up to the present Web. A specific URL would specify what its document type was (article, index, gallery, discussion, multimedia, catalogue, application, custom, ...), and the person accessing the site could define what style they prefer for viewing the site. The templates would be fairly tightly defined, but would allow for limited branding elements, think of how pre-printed paper letterhead looked back in the day.

Another Web element that might be interesting to disrupt would be search. If websites themselves could provide indices to their content, crawlers would not be necessary (though audits to guard against keyword stuffing or other deceptive practices would remain). Indices could be distributed in a mechanism similar to DNS. Reputation and ranking would remain issues (though those might be achieved through third-party or federated services). But much of the present infrastructure that's presently required to build search engines, and the entrenched incumbent advantage, would be broken.

I'd love to see a new set of Web architects play with ideas such as this, and am trying to do so myself.

just use Falkon. browser by the KDE team. free, open etc. https://www.falkon.org/

Go for it. Sounds like a good use of time compared to listening to the radio or browsing tech news.

I'm in a similar place, where for the most part I'd rather be in a kitchen washing dishes for minimum wage than managed by a tech guy (and have primarily been out of tech for 6 years, occasionally freelancing). Although I am strongly called to build my vision of tech and continue to do so on my own time.

There are similar solo "built by a crazy person" projects out there like TempleOS where you could find synergy. I'm another "crazy person" who's building a stack with early-2000s web tech ("web 1.5"), so if you look hard there's even synergy there... support 2005-era web and I'm on board.

Cheers.

Tens of people seems generous

Same. I even use gmail in the browser on Firefox/iOS, and all it does is give me a splash page saying the app is better, then I click through.

Not in Gmail, but for example, in Google Cloud webui.

You never thought about the why and what for?

It really isn't.

> Writing and maintaining a state-of-art browser engine is hard

I know. And the KDE team would know too, having tried (and succeeded at the time! but could not keep up)

> unless you don't care about people actually using your browser

Why so dismissive?

Blink is not the only existing browser engine that works with most of the Web.

I'm not arguing for Falkon authors to build their own engine. That would be fantastic, but I understand they don't. I also understand their choice to use what Qt provides. Embedding another browser engine would be less straightforward or even outright painful / impossible (Gecko). Still, I don't want to strengthen Chrome dominance on the Web.

It doesn't matter whether the author's browser succeeds or not.

What matters is that this project gives the author a meaningful sense of agency in the world and we all need that in order to be psychologically healthy.

Humans were not meant to sit passively consuming media that makes us feel bad all day. We're animals and animals live by doing.

For those interested in trying this on their own as a learning experience, I've really loved the early material in this book: "Web Browser Engineering." (https://browser.engineering/) It was posted here a while back.

I want a version of this:

Firefox with private mode and uBlock on but saving history (so _I_ can go back to something I browsed) with an option for saving cookies on a particular site (but maybe use a separate container so I'm not tracked across these). With an additional button to clear all random cookies now

I wish more people would try this.

People who can't or are unwilling to do something themselves often try to project that unwillingness onto other people.

Good Luck!

You can boycott the bad companies. Spread the word about how they are bad so others can stop using them as well. We live in a world where anyone can communicate with anyone else on the planet. The power is in the hands of the many.

When I got fed up with it, I stopped using a browser.^1 Never looked back.

1. Of the advertising-supported, graphical variety.

We'll likely always have to use advertising-supported browsers for some things.^2 But certainly not all, or even most, things.

2. It would be really cool if this sad state of affairs changed before I'm dead, but I have no expectations. I would love to be able to use something "like" an API for internet banking and online shopping, instead of an advertising-sponsored browser.

I feel like this is something which a government should be developing and funding.

I understand that governments like to make rules and hope that the market responds with something that matches their vision, but as the EU demonstrated this just ends up with websites with user-consent pop-ups. A worse experience for everyone, and bad actors just find a way around it anyway.

Rather than relying on legislation to play whack-a-mole, why not just build something for everyone with tax dollars and open source.

Is there some reason why browsers aren't split into a million pieces and worked on seperately? Most browser projects I see are tackled like giant monolithic projects.

For example, CSS seems like a constraint solver that has nothing to do with parsing HTTP requests. Even parsing CSS itself has nothing to do with laying it out. Rendering is another step altogether that also has very little to do with anything. I can see these things being split into standalone and testable libraries with limited scope. Maybe I'm wrong though. I often am.

A "browser" seems like an "operating system". There is so much going on and I can see how a "browser" could be made out of myriad existing components in various amounts. A bit like the various Linux distributions and their particular choices of underlying tools and frameworks. You don't do it all at once, you got a million separate small chunks working together (if you're a Unix kind of guy..)

Maybe we should all work on creating as much small composable components as possible and then let the other guys slap a nice UI on top of it and put a bow on it. This is probably not practical, but a guy can dream.

I'm surprised at the amount of attention that a threat to write a browser has gotten here. There is 0.000000% chance that anyone will write a new web browser that can reasonably render current web pages without about $10M/yr of budget or volunteer effort, probably much more. You can say this is just being dismissive or pessimistic or whatever, but it's no different than someone saying they are going to build their own Space Shuttle, and the progress so far is a partial sketch of one of the tires.

In about 3 or 4 weeks the author here has managed to implement about 2% of a very poor parser for html. Parsing HTML is probably one of the least labor intensive parts of writing a browser and is a task that would be suitable as a project for an undergraduate student majoring in CS as a project for a class. Writing a javascript engine that is fast enough to allow current web pages to work is not even comparable in complexity and effort. Projecting to what it would take to build a very poor browser, we are looking at about 150 years or more.

If I had to redesign the modern web and ditch the legacy requirements, I’d pick a design with a series of simple docs and vm’s that would look like a squeak/smalltalk environment. Directories and search engines would just be a plugin tool to index what you can retrieve with your permission levels. And session management would allow coexisting environments to run side by side like ide’s

I guess what I'm confused by here is that the author seemed to recognize that news is inherently biased to be negative, that negative (general) news was never actually important or actionable to them, and that getting that news was just bad for them overall.

Then he got annoyed at tech news being bad, and came to the opposite conclusion. That all the negative tech news must be correct, and that they have the agency to change it.

But the reality is the opposite. Just like for general news, tech newd has a bias for the negative, and the reality is a lot better than you'd think if you just take the HN frontpage at face value.

The scenarios that motivate the author are so improbable that they border on absurd. And if they were to happen, the plan of writing your own browser wouldn't help anyway, or at least would not be anywhere near the best way of addressing it.

Want to write your own browser as a hobby? Great, go for it! But the justification isn't even internally confident.

Anyone doing this for cars? I feel cars are much worse than browsers in terms of screwing their buyers and much easier to topple.

Way to go. Godspeed.

I, too, am fed up with It! That's why I'm implementing an open-source ASN.1 library.

...On the way to writing a new hypertext protocol that uses ASN.1, of course.

...Which I can then write a browser for.

I, too, am not sure if this will ever finish, but I figure that at least, at some point, I will definitely be able to tell you why ASN.1 didn't win the protocol wars. :-)

The first part of this blog really resonated with me. Stuff is bad, fascism is on the rise, mega corporations control more and more wealth, those two things feed on each other, book banning in the US, war in Ukraine, my country is burning due to global warming, ... and all of this bad news damages the psyche.

I've been thinking of how I can help, that I should stop reading so much negative news, for my own mental health, and do something to move the needle back towards the positive.

"This time, I DO have some influence because I am a programmer. I can develop software ..."

It was jarring to see the blogger go to "I'm going to write a browser". This isn't helping, this isn't even "a way to push back" ... it's a hobby project, like woodworking. Pushing back would be helping out with one of the open source browser projects already out there or something else productive.

I mean, "you do you", it was just quite the twist.

> "I’m not an activist fighting for a better world or anything. But I get angry about the corporations screwing over people, software, planets. I’m not in a position to solve that."

You ARE in a position to make the world a little bit better. We all are. Start with simple things, like thinking about using cc-by license and think about why, when and what FOSS license to use. A FOSS license is just a tool but make it work for your goal(s)... And why not just ask your local government to use FOSS software on websites to be a bit more compatible with every home baked browser. Your privacy and freedom is limited year after year.

The link to your Github project in the article leads to somewhere else

Go for it! I think you will learn a lot, and it seems like your expectations are calibrated well.

Source: wrote a browser[1]. I enjoyed it and learned a lot. My motivation for starting it is that the open source browser landscape was bleak at the time. There was khtml, which seemed fairly solid, but that was KDE and I was firmly in the Gnome camp. Mine was very fast and lean, but wouldn't scale to embrace what's now considered the basics of a web browser: JS and CSS. Motivation tailed off sharply when Mozilla announced they would be open sourcing their browser. If I had known it would take them over 4 years to ship Firefox 1.0, I might have kept at it.

Gambatte!

[1]: https://levien.com/gzilla/

I love this kind of naivitee! This is the magic stuff that either takes one to greatness or just a step further into nowhere, but it’s the most fun way IMHO to approach things. Go on OP, write a beautiful story for all of us!

You do it right man. I sit in the same boat as big corporations just start to go on my nerves. I use Linux, try to stay away from most google software but use a chinese mobile so I do not need to be logged in any Google or Microsoft based store. It is tough but I think you go the right way. If I can be of any help let me know. Maybe you create a github (bah Microsoft) or similar repository where people can help you testing and/or developing. You don't have to start completely from scratch as there are some engines available. I wish you luck and dedication.

It's good to have a project, more power to you if you can make a useful browser on your own these days. It's like an advanced OS at this point.

Didn't like the doomerism about the news in this post. Yea, you as an individual can't do much, but joining a social movement organization would make you not an individual. I think a lot of people here just don't want to take that step or were never invited to do so and so write off things as impossible.

mozilla servo all over again?

Rumors (I did not check rust syntax specifications) say rust syntax lost itself like c++ (and what is happening with latest C syntax additions from ISO).

Is this true?

Because, personally, I am not interested in a browser written with a beyond sanity complex language syntax. I would have prefered a web engine written in plain and simple C (without the latest tantrums from ISO). Without bazillions of deps, SDK included.

That said, a modern web engine, beyond a javascript engine (for instance quickjs), is tens of millions of lines of code (Big Tech made sure Small Tech could not reasonably follow).

In the end, in my opinion, the most reasonable option is to go noscript/basic (x)html where it does a good enough job, bridges with super simple protocols you can navigate to with URIs (for instance irc://... mailto://...), and really in the worst case scenario a semantically optimized network protocol (the leanest possible) with a rich GUI client (for instance sliding maps, even though you can browse maps ok with a noscript/basic (x)html browsers).

I want to write a browser that unifies all webpages to one single layout, one single interface that looks the same on all webpages. Minimal design, high performance, low latency, the only thing that changes is the images, text and videos on each web page Everything else is filtered and removed No animation No fancy graphics No emojis No additional symbols other than the text carrying the content.

Thank you for speaking to what I have been feeling for a long time. I have been recently thinking about starting some kind of organization or initiative to push back against the anti-competitive behavior behavior that are rampant these days. It would be similar to EFF, but more about raising awareness in corporations by sending letters, and talking to employees as an organized effort.

> I’m not an idyllic person. I’m not an activist fighting for a better world or anything.

Genuinely, why not? What’s wrong with being an activist fighting for a better world?

I see your kind of disclaimer all the time and I’m baffled that anyone would actively reject this nearly universally accepted pro-social perspective.

Someone help me here

Is the risk of being seen as Naiive really so strong as to overtake the highest ideal of humanity?

Most countries do better every year. There are only a small number of countries that are worse off than, say, five years ago.

Especially giants like China and India are doing very well.

See: https://ourworldindata.org/human-development-index

most of the bad news of companies is driven by the profit motive. that's where the connection to ads is. they are also driven by the profit motive. (there is also news about companies pushing more ads)

the conclusion of this author is that the profit motive will drive companies to build ads into browsers. (microsoft is already putting ads into windows. browsers will be next)

That plus:

- Impossible to repair without going to the vendor.

- subscription for seat warmers and "full accelerating curve" (https://www.theverge.com/2022/11/23/23474969/mercedes-car-su...)

- random feature disabling (https://www.wired.com/story/fight-right-repair-cars-turns-ug...)

Let us take as granted it that "the number of violent conflicts is the highest since World War Two". Let us assume that this is accurate.

Nevertheless, these statistics are globally better than any previous century: rate of people dying in war, poverty, child mortality, child education, female education, access to clean water, access to food, access to medical care, economy, life expectancy, crime rate, number living under democracy.

It will always be possible to point at one or another thing and truthfully say that thing is worse than it has been for a long time.

However, as bad as things are, as truly grim and miserable and precarious the lives of some people, we do indeed live in times of unprecedented peace and prosperity.

1. I didn't suggest "outright ignoring reality"

2. Yes, violent conflicts are increasing. That does not mean we are not living in historically the most peaceful times. This is indeed a case of statistics, while not outright lying, being selectively singled out to present a narrative. How many people are dying in these violent conflicts? How many people are being displaced? Child mortality, war crimes being committed, lasting negative economic impact, negative impact on international relations, etc, etc, etc, are all factors in whether or not the world is becoming more peaceful or not, and the overwhelming majority of those factors are improving.

3. "things are better than ever and always getting better is also not a fact as you say." - I challenge you, in your own time, to disprove that statement. That exercise does wonders for a persons outlook on life.

The Deputy Secretary-General of the UN might be one of the most news-brainwashed people on earth. The saviour complex you need to have to even join the UN is insane. We should not be taking advice from these out of touch billionaires.

That press release is some fear-based BS.

Care to share specific details?

Curl? wget? netcat? Terminal-mode browser?

(I use all of the above myself. Many websites ... behave poorly with them, though they can still be useful.)

Eastern europe and the US make up a mere ~8% of the worlds population

I strongly recommend you read the book Factfulness by Hans Rosling. It’s clear you feel strongly about these issues but have been misled by profit seeking media into doomscrolling news meant to keep you engaged at the cost of your mental wellbeing. This book will help you recalibrate your worldview.

Clearly, too much mainstream. About democracy, true. They could also be eroding private property. Besides those two things, everything is better: trchnology, better medicine, access to good education, longer livlfe expectation...

I hope you are not (I saw this a loooot) one of those people that to correct democracy things we should tell people what to do, think or say or not or what to love, hate, like, dislike or how to behave. I say this because I lived in Asia for 12 years and I see the westerner world collapsing on those premises.

Greetings.

US makes up a mere ~4% of the worlds population

Yes climate change and mass extinction is bad. Yes society will survive. No, climate change does not mean that things are getting worse.

> > and it's only getting better day by day.

This is the part I'm disputing. You are moving the goal posts to "generations" ...

There is a fair amount of evidence that we are currently in a down swing in some key areas.

> The world is actively combating and will fix climate change.

At this point, that's faith :-).

Look at the numbers: CO2 emissions are increasing every year. Climate change is happening faster than the models predicted. Biodiversity is dropping, we are living a mass extinction right now, it is an observed fact and not a prediction. Fossil fuels are not unlimited, we passed the peak of conventional oil production in 2008, the general peak is expected soon, same for LNG. We don't have any sustainable energy that can remotely replace fossil fuels.

The one thing that is pretty clear is that we must consume less. But we are not doing that, instead we are consuming more every year. Hiding behind some sort of "others will fix it, I have faith" Silicon Valley religion, or something.

Read. The. Article.

> I think this project will not result in an actual browser.

Maybe this project shouldn't result in a graphical browser. I still think there's value in text-based browsing, in the line of Lynx. I would think there's a higher probability of creating something really useful if you were to go that route.

I just want to thank you for sharing your politics views. These are the same ideas and views, as an American, that I’m struggling with. It helps so much to know other people in other parts of the world are going through the same thing and have the same thoughts and ideas.

go for it!

The hardest part of a browser is rendering. The rest of it is easy, relatively speaking

The big problem with browsers is state. Specifically, the web page has an internal state that the developer is tracking, and the browser UI has a state that the browser is tracking. How do you keep those two in sync?

The current model is based on the sort of hack that Netscape did way back when. There are probably other ways to do it.

Specifically, the problem is: how do you give something access to a representation of the visual state (the DOM) while hiding your implementation of it...in a secure manner?

And how do you allow that client (the website) write access to that DOM representation securely and safely?

I like your list of 100 opinions. Of course, not absolutely all of them ;-) but the effort you took to compile the list. Lots of things to think about.

You could do something back then, and it’s the same only thing you can do now: work with others. Drop the egocentrism, because collective action is the only way out, and it is possible.

OP you go write that code and profile and make sure you get some real world feedback on it. I hope it DOES result in a real functional browser.

Screw everyone else, it looks interesting and fun. I'll be following your RSS updates. Good luck!

love the name ... needs a 1980's Commodore BASIC visual :-)

Shut up and hack!

Respect

I would say though that you have to pick your battles. On your path you might vanquish shitty web sites but just not being able to read them, but I think a real victory against enshittification involves breaking away from the web user interface which is some of why there has been so much bot blocking lately. It's not that people are going train an LLM on a shitty site but that you're going to get the LLM to read the shitty site for you, that is the real death of shitty sites. (e.g. agent posts your question to Reddit on your behalf, picks out and summarizes the best answer; instead of looking at craigslist every day for farm animals you get notified of the good ones; instead of seeing an ad for a Laurie Anderson concert on the bus when the tickets are all sold out the system tells you the moment the tickets go on sale...) But who knows? An HTML 5 parser is a good place to start towards that end.

I had similar motivations for developing my smart RSS reader and intelligent agent YOShInOn which has gotten me to work on filtering and "deshittification" right away as opposed to face a 1500 man x year project before getting to the stuff that matters to me.

For me the path not taken was developing a "reader" that works like archive.today, there was a summmer in the pandemic when my son and I were driving back and forth to Buffalo a lot and talking about missile defense systems and I did a "spike" project to develop something that would archive, process and deshittify web pages and I came to the conclusion that it was a long road to make something that was better instead of worse (every archiver I've seen is way slower than just browsing the page and sorry "slow X" advocates, slow = bad)

This winter I turned to the problem of end-to-end active learning for selecting articles and had great success as it. Sometime it sends me to archive.today but the goddamn CAPTCHAs never end when I use it from home so the question of the archiver/reader has come back but now that the thing has a workflow engine and could archive articles before I look at them the archiver/reader looks appealing if I can find something open source that is good enough.

There's a commercial product

https://www.getupnext.com/

which took the opposite approach of developing a reader and now they are working on recommendations. That's a path that makes sense because the reader cleans up the text which in principle is good for NLP but for what I am doing I think the signals in dirty text have enough value that I am doing just fine. They've got the issue that they're very unlikely to really develop content-based recommendation for a multi-user system because collaborative filter is such a good shortcut, also YOShInOn uses a Tik Tok style interface which for some reason has hardly been tried despite being a hit with Stumbleupon, this interface collects good negative samples and gets great results with a simple classifer... if you get a few 1000 judgements.*

What I've found when you become a nuisance but read that as successful, you'll find all sorts of problems suddenly plague you in your off line world.

So I wish you the best of the luck, but keep an eye out, the world but in my experience the US and UK, is nothing like its "public persona".

Idea -- pump the whatwg specs through chatgpt to generate the Rust code.

This is great. Would you be interested to add support for a custom blockchain-based DNS instead of the current centralized hierarchical mess? I've been thinking about this a lot and it's clear that a multi-blockchain-based solution would be perfect for this (and far simpler than the current system). Each blockchain could have its own top level domain. People would buy domains using the blockchain's own tokens on the blockchain itself. The browser would just discover its own peers and query any node in the blockchain to convert domain names into IP addresses. The browser could check the results against known public keys on that chain to verify that the record is legitimate.

Blockchains are great for this use case as they are infinitely scalable in terms of reads (since you can always add more nodes which are exact cryptographically-verifiable replicas). While blockchains are constrained in terms of writes, this problem is dealt with via a transaction fee which increases with demand so it is naturally DoS-resistant and doesn't require special skills or trusted parties to operate.

Anyway, the current centralized DNS system of hierarchical resolvers is also constrained in terms of writes but it feels ad-hoc with many centralized entities which are given a substantial amount of trust.

It is terrifying to me how strong the public desire for retribution is. Despite over five years of conflict, the history books show that there will still many who wanted more bloodshed to 'get even'. As what might be considered an even greater tragedy, there were loyal members of the various French resistance groups who were summarily executed for their 'involvement' in Nazi or Vichy institutions, when in fact they were spying for the French against those very institutions.

from the "assume ignorance before maliciousness" standpoint, a lot of the impression you speak of is just a fact of us humans reporting on the (perceived) more critical news, (perhaps also maybe the more profitable news) - but is more (in my opinion) a natural result of the news having limited time/bandwidth to report on a near-infinite content production source (the world)

A side effect (and the end result) is that we all end up having poor [Bayesian priors](https://en.m.wikipedia.org/wiki/Prior_probability) in our heads - this disproportionate reporting practice at large impacts our understanding of the world, as you describe. And I bet there's a feedback loop: the more skewed our model of the world, the more we will produce/consume skewed content.

I think this is not something you can ever truly do completely away with because of the many factors that it arises from, not limited to our psychological biases that tend us towards liking or interacting with certain form of content more than others... but certainly there should be more controls.

There's a positive aspect to employing police officers to be present even when there is no tangible risk of crime or danger.

Where I live, it's extremely unusual to ever see a 'bobby on the beat', so when I recently visited a city with thirty or more police officers with riot vans congregated at the docks, ironically I felt slightly uncomfortable - what mortal danger could be present around the corner to justify so much police attention? It turned out to just be a big football match, so there was not such a huge danger after all and the police would no doubt be able to diffuse any disturbance immediately.

If there had been no police around, it would never have occurred to me that there was a risk, even though I would have been tangibly less safe without their presence.

> But I guess that’s just me rehashing what you said about reaching critical mass.

Yes. But one should remember that some services that are popular today started with a very small user base. Programming languages have to start from one user (starting from zero is orders of magnitude harder - if you are not the first regular user, your project will most likely fail).

One way to achieve that is to have a "killer feature" or maybe a unique combination of existing features. For instance, a subject related to browsers is discoverability and search. If you are going to compete with the WWW without funding or support from companies/institutions, there's little hope that someone will provide a search engine. This means that users will have to pay for that service with their time (that's how Wikipedia emerged). One could think, for instance, about a content tagging system that users could share, to solve this problem.

There are HTTP to Gemini proxies/mirrors like gemini://gempaper.strangled.net/mirrorlist/ or Wikipedia mirror: gemini://vault.transjovian.org/en

Writing parsers honestly isn't that hard, especially when you have more or less step-by-step instructions for some of the formats, and test suites that independent 3rd parties have written that you can use to help verify your implementation.

Regardless, so what if it's "bad"? OP is using this as a learning experience too. If he wants to write parsers, and thinks that experience will be good for him, then he should write parsers.

I do agree that writing your own parsers isn't strictly necessary if you want to write an independent web renderer/browser. If I were tackling this project, I'd be more interested in and curious about how to write a renderer, and if there's a HTML or CSS parsing crate already written and available on crates.io, I'd probably use it.

Then again, writing a parser would also make me intimately familiar with HTML and CSS in ways that I'm probably not currently, and that might help inform how I build the renderer.

yes- which is usually fun and teaches you a lot!

In the hierarchy of "from scratch", it goes more like "buy fast car", "build fast car from available parts", "machine own parts from stock", "make own castings from scrap metal", "mine ore and smelt it".

Just because my hot rod isn't made from existing parts does it mean I'm mining ore.

Many awesome hot rod projects involve machining from stock or even casting from scrap. What OP is doing is on this same level. He's not doing atomic layer deposition or anything. Just writing code on top of an existing OS and using existing programming languages.

I don't think he plans to build his own processors and write it in machine code.

An unfortunately large subset of too-comfortable tech workers have decreed the right way is to: never reinvent the wheel, always import, and avoid 'the hard stuff'.

The sad bit? They're well compensated for this learned inability, so arguably they made the truly smart choice here.

To the blessed few who reject this apathy and continue building new innovations? I salute you o7

Of course nobody said that. Just like nobody should be saying anything to this guy about writing a browser.

Actually Linus himself proclaimed it as a hobby that would never be big and professional:

>I'm doing a (free) operating system (just a hobby, won't be big and professional like gnu) for 386(486) AT clones.

Making a browser is a big undertaking, and I would set the same level of ambition. And then I wouldn’t post it on a forum until it has reached a milestone. :-)

Funny that it's now unclear if people are talking about the network formerly known as Twitter, or a hypothetical.

Seems like it would be possible to write a formal proof that "X" is the worst possible brand. It's like he just walked into Walmart and grabbed https://www.walmart.com/ip/Letter-X-Branding-Iron/485919721

Do you hate specificly the comic, or that people often post it, where it is not contributing to the discussion?

"Goes without saying, its a browser, browsers are never finished. "

But they can be working enough, to be useful. Meaning they can display correctly the pages you want to see.

So the best independent effort of a working browser I know of, is ladybird. And Kling knows how to build browsers, so toying around with a browser from scratch on your own will be a learning experience. But I guess you might learn more, by contributing to a project that stands a real chance to become a working browser.

> What you end up with is another Gemini Protocoll that a few nerds use and the rest of the world doesn't care about.

Reader view is already half way there, and it is as mainstream as it can be, as it comes on all iPhones and Macs. But you're right, my idea doesn't have much to do with the underlying protocols.

Speaking of branding, marketing and UX design, I would say that most businesses have already completely given up on that, and are content to sell their stuff online through portals that are completely controlled by third parties such as Amazon, Booking.com, Doordash, Instagram, etc. Very few make a decent effort to have a website where their customers can easily purchase their products, even though it isn't hard at all.

Could a user-centric browser extract the HTML results of everything made by javascript and present it in a standardized way? Because text, images and videos are already solved problems with reader view. The interesting part would be if HTML input elements could also be thrown in, even though they're usually quite javascript connected. Any website more complicated than that is more of a web app, and should be rendered as the designers intend, in my opinion.

In the end, I think even a folder structure would be the right way for front-end as well as back-end for most websites. Here's your articles in HTML, here's your images, here's the product you can purchase as some kind of standardized json-object. Does this make sense?

Gemini can, in fact, serve up images, audio and even video. The specification disallows inline display of such types, but it can serve them fine.

IMO most of what people do on the web is either reading articles (these don't even need JS), watching videos (no advanced browser functionality needed either), or writing comments and messages (some rather simple JS if any).

I was interested in Servo, but its website is exasperating - any idea where I can find info on what HTML, CSS and JS standards it supports?

Quantum (mentioned by parent post) is what Firefox calls their Servo bindings.

I didn't flag/mod the comment, but I think it errs towards conspiracy theories.

I'm not American, and I understand the American government does evil things sometimes, but it also does a lot of good things for its citizens. It builds roads, keeps the skies clear, funds education, health systems, sues tech giants when they go too far, tries to manage the speculation/money markets, etc. To say "undeniably often working against the best interests of its citizens" is very hyperbolic. Most would probably argue that the government does not do enough, even conservatives (ex: controlling abortion or bathrooms, not that I agree).

The news are also not there to demoralize you, not as a primary objective. Anxiety and negative news do sell ads. I find public radio such as NPR do a wonderful job of balancing point of views, giving the facts, and provide some hope by explaining the underlying issues and possible solutions. Where I live, public radio is the only non-ragebait media that constantly puts pressures on whatever government is in power, and explains various points of views (while admittedly being more centrist editorially, angering both hard-left and hard-right).

I totally agree however with doing whatever it is you are good at, and that's a way to work towards improving the world.

Even in that country - if you’re a woman you need to know if you still have rights regarding bodily autonomy (e.g. abortion). If you’re queer you need to know if you’re allowed to dress a certain way (anti-drag laws) or express public affection, without being arrested.

I am guilty of both, naturally :)

> To complete the analogy, I guess you’d expect to see this browser come out of former Safari interns/engineers at Apple.

Sounds like Andreas Kling: https://www.techgamingreport.com/new-ladybird-web-browser-wh...

People interested in building browsers should look into WebKit. You can use that for HTML/CSS/JS rendering and save a lot of time/effort.

> The basic thing I am saying is that if you call it a web browser and use the same core protocols etc., you have a practically infinite (for an individual) set of requirements to implement and will never be truly compatible

But what the parent is saying is you don't have to be compatible with 100% of websites. If you build a browser with good HTML/CSS/Javascript handling, you'll be compatible with the large majority of the web. Yes, there are other specs that are in use by a tiny portion of the web. But your browser can still exist if it doesn't support them. Users of that website will know or find out that they can't use your browser for that, but for every other user that browser could be exactly what they need.

> without having to worry about a fuzzy and ever-expanding definition of what you have to support.

It's only ever-expanding if you're trying to keep up with the Jones's. You can define a subset of the web you want to support and build a clean, small browser that supports that.

I would not want a simple browser to replace gemini, but I would be happy to use a simple web browser in addition to gemini. It would be nice to only have to fire up a big bloated real browser for doing banking and such things.

Isn't it called `netsurf` already?

I'm not upset at the OP, but I wouldn't say this is a golden rule. Some people make some really atrocious shit. People should, in general, be at least a little critical of what they're making.

it might be easier, but that doesn't mean all criticism should be ignored.

In the spirit of cliched reductive "feel good" statements of this whole comment section, ignoring criticism is also easier than building something that ends up to be extremely subpar.

He doesn’t, if he’s honestly ok with that.

But you only have so much time, why not pick a goal that is achievable? Seems like a waste.

I've literally only had success with software I built to satisfy my own curiosities and needs. Not all of it took off, granted, but I think that's fine. Not everything needs to succeed.

Current system does have many nodes in a hierarchy but they are operated by trusted entities and there is not much fluidity. Also nobody owns their domain names; just renting from the centralized providers and you just have to trust them with everything.

With a blockchain, you would provably own your domain name and nobody would be able to change your 'zone file' aside from you since they wouldn't know your private key. I mean it would be physically impossible even if they wanted to do it illegally (or if the TLD operated under a crooked government).

BTH, this has already been done by Unlimited Domains (.eth TLD) but for some reason it still hasn't been integrated into major browsers (aside from Brave which has it disabled by default though it definitely works).

Yet Bitcoin's market cap is $500 Billion and has a trading volume is over $10 billion per day and has been outperforming the markets for over a decade.

If the financial system is fine, then how is it possible that a 'scam' is worth $500 billion? What kind of financial system allows such scam to exist for over a decade and keeps attracting investors year after year?

If the financial system is not fine, then how could you say a blockchain like Bitcoin is a solution looking for a problem?

That position doesn't make sense from any angle.

And anyway, I've explained clearly what is the advantage of blockchain:

- It's a system which supports unlimited reads at scale.

- Spam-resistant when it comes to writes.

- Its data can be verified independently so there is no need to trust service operators.

Those are the most important characteristics of a DNS. Sadly the current system is missing the third point and that's why nobody truly owns their domain names. Everyone is merely renting from a TLD provider and hope that the laws of whatever country it's in are going to respect your claims.

I despise cross-platform GUIs as a concept, but if you must make one, there's plenty of options usable from most languages. Java has Swing, C++ has Qt, there's things like Flutter and WxWidgets and a dozen different GUI libraries for Rust. Heck, you can even build GUIs in Python with tkinter. Every single one of these options is much better suited for an app than the web stack.

My main gripe with using the web stack for building apps is that, unlike proper GUI framework, your platform is not made out of the same kind of code that you're writing. For example, there's a number of predefined layout algorithms that you can choose from and then control some of their parameters, but you can't write your own layout algorithm. You can kinda draw your custom controls, but both <canvas> and SVG are not without their own shortcomings compared to real GUI frameworks. A web browser is a glorified scriptable word processor so it's not surprising that it acts like one.

Yes, there are "paint worklets" and "layout worklets" as proposals. That kinda makes it better but still feels like an ugly workaround. The declarativeness of HTML and CSS, extremely awkward for building app-like UIs with, is also never going anywhere.

If a "hidden empty div" is all it takes to break your browser's performance, I'm sorry, that's just not impressive.

Thankfully you don't have any idea of my (current) location.

I could be anywhere in the world. Even nextdoor to you ;)

You and GP still would be deluded however.

I've yet to see someone take a PWA seriously.

"Not all those who wander are lost." -Tolkein

You and I obviously have different value systems so this thread will never coalesce.

Enjoy your travels towards your goals.

"If you mean by setting `position: absolute` and then using left/top/right/bottom to position your element, that won't work."

Erm, but it does? You can listen to resize events and you can get the exact width and height values of all the elements in the DOM. I am using that technic since I switched to the web (when box layout did not exist yet).

The only problem with this is, that you can create performance bottlenecks easily, as reading clientWidth might trigger a forced reflow. So it might not be easy at first to get everything right, but it is totally possible.

Also you can just use a different GUI framework or build your own.

"Not everyone's goal is to simply have as many users as possible."

Mine neither, but those who do (the majority) made sure, the web is suitable to be used by even the most tech illiterate people. My mother can surf the web to get mostly what she wants - but she cannot use a file browser, nor does she really know what a file is (and she does not want to know).

"> The app you make, you can ship by simply sending a link.

Depends on what it does."

Sure, if you want native file system access, you would be breaking the sandbox. And yes, this is my biggest pain point with the web as well - that it often cannot decide whether it serves untrusted websites or enables app developement where the user trust the app. For example there is no (easy) way to get the pixel data from html elements, because of security. Which makes sense for malicious add filled websites, but not when developing something real, where the user could give permission. Here the focus on the dumbest users who just all allow everything, so the browsers have to protect them and allow no one to use proper permissions is really annoying. Because yes, I also would just like to have file access and other things permissions don't allow.

IMO, not being able to recover a lost private key is an acceptable tradeoff for fully owning your own domain in perpetuity. If we encounter a de-globalization event and international law starts to break apart, we would see a lot of value in not being beholden to foreign entities operating under foreign laws.

People are innocent until proven guilty. Corporations are driven by greed and profit until proven otherwise.

I would be skeptical too if it was just a one time thing, but it wasn't. Around the same time as that HN comment, YouTube rolled out a redesign of their frontend that used the Polymer framework. At the time, Polymer depended on a deprecated web standard that was implemented in Blink and WebKit and used a polyfill elsewhere. This completely killed performance on Firefox and Edge. That isn't unsourced or anonymous information like the invisible div. You can search back and find people complaining on multiple forums that YouTube started running at a crawl out of nowhere and ID'ing Polymer as the culprit [1, 2, 3].

[1] https://hub.packtpub.com/youtubes-polymer-redesign-doesnt-li...

[2] https://news.ycombinator.com/item?id=17612139

[3] https://www.reddit.com/r/programming/comments/91i0mc/youtube...

I have no idea if the story is true or not, hence the "if true" clause in my post. I was just pointing out that focusing on the technical aspect is missing the point of the story... the degradation could have been triggered by any number of things, the point of the story was the allegation that Google purposely triggered it and disingenuously used that degradation to claim their browser was faster.

This isn't a court room.

Well, you'll find it hard to argue with people if you let your philosophical beliefs get in the way of finding common ground. Many people don't agree that "innocent until proven guilty" does not (philosophically) apply to corporations.

I think a lot of distaste towards GPL comes from devs with corporate day jobs who feel frustrated that anything GPL licensed is functionally off-limits at work because there’s no way legal would approve of using it, and certainly would not approve of open sourcing whatever project is built with the GPL thing. I can see where such an individual might think to themselves, “I don’t want to put other devs in my position in the same pickle, so I’m licensing my thing as MIT/BSD”.

As a developer my problem with the GPL is that it's less free than other licenses. It's an interesting legal hack, but it can only succeed in its goals by boiling the ocean. I just want to write things, and sometimes making money involves not publishing all of your source code. Having to choose between options and idealism limits the popularity of GPL projects.

  It's not for you to exploit to make a buck

  What you are describing must have been written by someone who does not believe in software freedoms.

  I've made lots of money and I publish every piece of code I've ever written to do it.

I think opinions around software licenses are limited to us hackers, and the actively code-publishing subset of us, moreover.

From my POV the GPL certainly seems alive and well, and the ideology of the Free Software movement remains sound and strong.

What I'm alluding to is the wider public. In <podcast-i-dare-not-mention> we literally walked around town interviewing young and old. Soon as I can, I'll share the results in a HN submission.

Copyleft was a good idea, but its real world utility is a lot less than anyone would like.

I once inherited a GPL project, but I couldn't muster any community interest in working with it. Plugin authors would have been compelled to use GPL and no one seemed interested. We had to burn the project to the ground and start over with a new MIT licensed project. Now there's a flourishing community of plugin authors working with it.

GPL being a viral license is the core problem, I think. I get the intent behind that clause, but developers tend to not enjoy being mandated to like that.

I don't think GPL is bad, but it's not exactly pleasant either.

I don't use the GPL (and never have) because, like another poster said, it is less free. When I put work out there for someone to do what they wish with it, I don't think I should then say "oh unless you do xyz, I don't approve of that". That strikes me as hypocritical, and so I don't do it. I don't harp on anyone who uses the GPL, they gotta do what they think is right. But for me, the GPL was never a good choice for software licensing.

And before anyone says "someone can make a closed fork of your code and then charge for it", I am aware of this. I also don't see the problem. My goals when sharing code are to share my work and help others by doing so. If some company uses my work to profit, I still have accomplished my goals. Moreover, others can still download my original code if they wish and reap the same benefits as the for-profit company did. This is a great outcome. Everyone benefits, nobody is harmed. While this scenario is often used as an argument that permissive licenses are bad, I think it is a very poor argument indeed.

I choose MIT as my default license, because I don't like it when the license is longer and/or harder to understand than the program itself. Simplicity is underrated in software development; there's so much you can do with less.

To me the biggest problem with GPL is compatibility. The majority of coding effort is commercial and will be for the foreseeable future. If the frameworks are GPL then the commercial interfaces will likely be incompatible with free software. Thankfully they are mostly BSD so commercial software works on free software systems and vice versa.

For user-facing software, GPL is fine, but fewer people actually want to write that.

I've seen too many cases, recent example comes to mind is apple switching from bash to zsh, that are done not for technical or even business reason but ideological.

I'm simply explaining the point of GPL. Being able to work around the point doesn't mean it's not the point.

Do you really think it's fine and good for Google to take work explicitly intended by its authors to preserve free computing and work around the licensing to make money without preserving the intent and purpose of its license?

Shaming would be much stronger. I even pointed out that participating in it is totally elective and I personally would not in many cases.

Framing it in terms of your own cost/benefit is missing the point, though, and there is a meaningful and valuable purpose to it outside of maximizing your personal benefit in the moment of choosing a particular library.

The FAQ is just the FSF's wishful thinking, and not a legal document.

If your program can fork and exec another GPL-ed program, and exchange complex data structures with it, that is neither here nor there.

All that matters is whether or not you are redistributing the program, and if so, whether that is in accordance with its license, which restricts the manners of redistribution.

If you are not shipping that program, but your own GPL-incompatible program needs it (your program cannot be used unless the user has an installation of the GPLed program) then you're likely in infringing waters, because it could be seen as redistribution anyway. Though it you aren't physically redistributing the GPLed program on the same medium, redistribution of that GPLed program, carried out by someone somehow, has to take place for your program to work. It's as if you are redistributing it "by proxy". If the user doesn't have that program, you have to instruct the user on how to obtain it, thereby turning that user into an agent of redistribution, acting on your behalf.

But if you ship another program which is drop-in compatible with that the GPLed one, which allows your main program to run, then in all likelihood the GPL doesn't have a leg to stand on against you.

If you don't redistribute a GPLed program, in any manner, whether directly or "by proxy", you cannot be infringing on its license.

A GPLed program being combined with yours by the user, in their installation, has nothing to do with you, if that combining isn't a condition of your program being able to usefully execute. That is not redistribution but use, which the GPL does not restrict.

For most people, a pile of software is not nearly enough. They need help of every sort to use it effectively.

Dual licensing becomes very difficult once there is more than one person writing code, and that includes submitting bug fixes. Once this happens you need to track down every contributor and have them sign licensing terms. It's long, difficult, expensive, and you might prefer to rewrite the whole thing again. https://en.wikipedia.org/wiki/Contributor_License_Agreement

The difference is that nobody is harmed in the scenario where someone makes a closed source product with permissively licensed code. The original project is still there and still just as usable. Everyone still has the same freedoms as they did previously.

The reason we have custom roms on android phones, is because of Linux's copyleft license: Phone vendors are required to publish their kernel modifications shipped with their devices. This already reduces e-waste by lowering the purchases of new phones, when old ones are perfectly functional. Thus, copyleft helped the environment

I don’t like hypothetical counterfactuals but of course it is your choice to use them. It is hard to meaningfully compare recent programs, rust or otherwise, with established behemoths like Linux, gcc, Emacs, etc, that did the heavy lifting and allowed the playing field to remain even over multiple decades. If gcc did not exist it is not clear at all how a hypothetical computerized world would develop.

The point of copyleft is to protect the continued existence and welfare of such a commons. I'm not even necessarily saying it is the only way to achieve a healthy free commons, I'm simply explaining what GPL is for.

That said, do you think it is in fact protecting the user from exploitation, or promoting the health of a free commons, to use "free" software in nonfree products which do not provide their source code to the user nor guarantee the user's ability to study, extend, and modify the software that they are using?

I think the argument for Free Software is at the very least strong and good. There can be other approaches to solving the problems FSF wants to solve, too, but it is a totally valid and non-debatably historically useful position at the very least.

> The GPL is a contact that takes away some rights of the people using it, and that makes everyone freer?

It is not that simple. The GPL "takes away" some rights of the distributor of the software (their ability to restrict redistribution for example), but grants rights to the end user (the ability to modify and redistribute) of the software.

It's all perspective. As the developer the GPL may feel limiting, but as a user it is liberating.

Unfortunately you've fallen victim to years of anti-GPL propaganda and FUD. A 10-liner GPL Python script is not going to "infect" an org with 20 million lines of closed source code. Suggesting that it's possible, even "in theory", is absurd.

Yes, you need to be careful when you integrate GPL and LGPL code into your proprietary code base, and in some cases you just cannot do it. But so what? You are not entitled to use of that code. If you can't use it because of its license, then that's life: move on and find something else that you can use, or write it yourself.

And you also seem to think that if you screw up, and do depend on a GPL library (for example) in a proprietary code base, then you're required to give away the source for your 20 million proprietary lines of code. But that's not the case. There are many different remedies possible. One (common) remedy might be to cease distribution of the infringing code, remove the GPL component, replace it with something else, and re-release.

I think there are very few copyright owners who release their code under the GPL who would immediately push for the remedy to be your company releasing all of their proprietary code under the GPL. At least certainly not at first contact, if the company is capable of acknowledging their mistake and working in good faith to correct it. And even if a copyright owner did decide to push for this, it is ultimately up to a court to decide, and it seems vanishingly unlikely that a court would order that remedy, especially if the company was offering less-extreme, good-faith options.

> It is trying to force my will on other people’s code

I just wanted to call this out specifically, because, again: absurd. Please look at it from the default state: you do not get to use someone else's code at all. That's what copyright law says, in the absence of another agreement. Even if you have the source code to my library (or whatever), you cannot use it in your product unless I have given you a license to do so.

If you are considering using someone else's code in your project, one of the first things you should do is look at the license. If you see GPL, and don't want to abide by the terms of that license, you move on and find something else. That code is not trying to "force its will" on you. You are free to use it and follow its licensing terms, or move on. You are no poorer because that code is there, as you are not entitled to use it without a license anyway.

  Unfortunately you've fallen victim to years of anti-GPL propaganda and FUD. A 10-liner GPL Python script is not going to "infect" an org with 20 million lines of closed source code. Suggesting that it's possible, even "in theory", is absurd.

IANAL but I don't think your example is accurate. Unless you expanded that 20 line python script into your 20 million line application, the GPL is not going to cross over and infect your other code. It would mean that if you are distributing that python script to users then you have to include the source for it (not usually a problem with python, but it is for other languages).

The GPL started when RMS wanted to fix his printer and Xerox said "No."

Ok, I also have no hard data at hand, but in my impression the vast majority of projects even started by individuals today do not use the GPL, but rather MIT or alike. Me too, because I do not like the concept of copyleft. I like the intention behind it, a world with open source as standard - but I just think the way towards this, is by convincing people step by step and proofing that they can also make money with open source and not trying to force them to open source.