This blogpost starts with me switching of my car radio, and ends with me writing a browser. There is some stuff in between as well.
Interesting take from the author; exactly the kind of thing that might start something big — or maybe it won’t, and that’s OK, too. Either way, I can appreciate the attitude!
(There’s also a discussion on the orange site)
The author is a bit early announcing their browser at the “proof of concept for an HTML parser” stage. Ladybird is probably a better “new” browser to follow. Originally written as a browser for SerenityOS, Ladybird is now actually getting funded to build the web engine into something practical.
It’s still in its early stages; layouts don’t look right, Javascript is very very slow, frequent crashes and errors, but on the other hand, it’s developing quite rapidly for a complete browser rewrite.
Thanks for pointing out Ladybird. It’s a pretty exciting project. But the author isn’t early in “announcing” anything. This isn’t a press release. He posted on his own blog about a pet project. That’s what the web is supposed to be. Not everything has to be for a big purpose or compete with everything else.
The author of the article can’t even be bothered to keep his server up-to-date (my first attempt at viewing the article bounced me with a warning that suggests he only has obsolete crypto protocols available for SSL—why bother with SSL at all, then?). He’s quite correct that this initiative is going to come to nothing.
There are currently only four web rendering engines that could be considered remotely usable as daily drivers: WebKit, its fork Blink, and Gecko, with its fork Goanna. WebKit and Blink both have major corporate backing (Apple and Google respectively). Gecko has the Mozilla Foundation paying the major bills. Even Pale Moon’s Goanna has multiple people working on it (and since it’s my daily driver, I know it has persistent issues with a few sites that have to be papered over with extensions). And the rendering engine is not the only thing you need for a browser, just the largest single part. A one-man project starting from scratch is not going to be viable in this day and age.
A one-man project starting from scratch is not going to be viable in this day and age.
It’s a pet project; it doesn’t need to be “viable”.
I think this attitude is part of the reason why we have so few browsers. Every time someone tries to start their own browser, even just for fun, a lot of the response is just bitching about how big and complex browsers are and how the effort to start a new one is wasted. It makes it so that people interested in writing their own browser (for fun or profit) are less likely to share about it and probably less likely to pursue it seriously
Not everyone is making something for you to consume.
He did say he was doing it to learn. Maybe when he’s done he will be able to appreciate what goes into making a viable browser.
Don’t forget the greatest rendering engine of all: Lynx
I concede the point, although Lynx is of limited usefulness on the Javascript-burdened modern Web.
I suppose that what bothers me about the original announcement is that it strikes me as something that shouldn’t have been announced until he had something with at least a Lynx level of viability. You don’t have to tell the world at large about every single hobby you take up.
Maybe I’m just too old and too private to understand People These Days and how they choose to go about their lives. 😅
I was only trying to make a joke. Your point definitely stands, and I agree with you.
Welcome to modern CV-padding.
Write a blog post about something (basic) you did.
Never mind that you just did it to have something to write about.
Go to conferences to talk about the blog post.
And the next time you change jobs you can pad your resume with all this stuff as if it makes you special.
A lot of companies actively encourage this behavior to market themselves as to better attract candidates.
And then a lot of companies indirectly encourage it through both their hiring process and possibly even their job ads.
Now, don’t take this the wrong way; there’s plenty of good talks at most conferences. There are people blogging about worthwhile software projects too, but there is a high volume of low effort content which really doesn’t add anything.
I’ve even been on both sides of the argument I’m making. Stuck listening to someone who doesn’t really know the topic, and stuck giving a talk about something I don’t really know enough about.
Oh my… Lynx really brings me back. I used to slack off at work by having an ssh connection to my home computer and browsing through Lynx (or was it Links?)
Never really got into Linux beyond basic use, but I was always proud at myself for managing to get this working :)
For a second I thought I was reading a comment on orangesite by mistake.
Dumb question, will it be possible to create a browser that will trick the shitty new web into thinking you’re using an approved non-ad-blocking-browser while you still block the malware/ads?
Building adblock into the browser could enable better countermeasures for adblock detection, but uBlock Origin’s filters usually work fine in my experience. Hiding that adblock is being used is essentially just an arms race between adblock detectors and ad blockers.
I believe they’re asking about the DRM enabled browser tech that Google just rolled out and is pushing webmasters to support.
Ah, my bad. Bypassing such integrity checks should still be doable, either by reverse engineering and spoofing the communications between the browser and Google, or by modifying a “trusted browser” in a way that keeps it from detecting such alterations. It might not be very reliable though, as the internals could be changed arbitrarily with each update, and old versions blocked in the name of security.
I hope you are right about this moving forward. My experience so far has been the same as yours (re: the arms race, uBlock working well). My understanding is Google is being super aggressive about this and they are not dumb, they know about uBlock, Pi Hole, and other ad blockers that people use. They want to make it impossible to use them by basically profiling your browser and its extensions. However, that could just kick off a variation on the arms race. A lot of things about their attitude makes me angry, but they are just ignoring the fact that advertisements are a vector for malware and don’t really care if they help spread malware as long as they are getting that sweet revenue.
What we need isn’t browsers. What we need is an universal way to write extensions cross-browser.
Browsers themselves are easy to make. The problem is convincing extension devs to work with yet another codebase.
E: Think of it this way. There’s a lot of open source browsers out there.
Are you using any of them? Probably not.
Would you use one if it doesn’t have for example Bitwarden, Ublock Origin, Sponsorblock, and such mandatory extensions?
Users follow extensions and ease of use; not what’s good for them.
E2: A good project would be a builder extension for VSC for example, which compiles to all supported browsers.
Browser devs would then contribute to said extension via native-made plugins.
Cooperation of two fronts.
Browsers themselves are easy to make.
What makes you say this? There are only 2 fully featured browsers, Chromium and Firefox. Anything else isn’t a viable alternative even if we compare stock to stock.
There’s a lot of open source browsers out there. Would you use one if it doesn’t have […] mandatory extensions?
There are literally only chromium-based browsers and Firefox (and its forks) with any meaningful market share. Developing a new browser engine is extremely complicated and time consuming, so there really is no danger of having “too many” browsers. And of course all browsers based on chromium (Google Chrome, Edge, Vivaldi, Brave, …) support the same set of extensions, because they use the same engine. So extension compatibility is also not a problem.
Supporting the gazillion ever-changing web technologies and standards and layout systems for a completly new browser is a problem though.
What we need is an universal way to write extensions cross-browser.
WebExtensions exist for this very reason. Chromium (and derivatives), Firefox, and Safari all use them.
browsers themselves are easy to make
That’s … a patently false statement.
They are among the most complex, difficult, resource hungry pieces of software out there along with actual operating systems.
There’s a lot of open source browsers out there. Are you using them? Probably not
This is also essentially misinformation. I’m sure none of us have heard of Firefox before, or Chromium. Sure Chrome (closed source) is what most people use, but Firefox isn’t exactly some esoteric browser.
This is not a standard. In this case it’s desirable to have more options.
