Can we really call edge a major browser in this context when its another chromium based browser? I'm sure there's more to it than slapping an edge mask on chromium but still.
Google Chrome, Microsoft Edge, Mozilla Firefox, and Apple’s Safari
Can we really call edge a major browser in this context when its another chromium based browser? I'm sure there's more to it than slapping an edge mask on chromium but still.
As a Norwegian, I'm a bit sad that Opera is no longer considered a major web browser. It's even been growing steadily in popularity for years now, and has almost caught up with Firefox in market...
As a Norwegian, I'm a bit sad that Opera is no longer considered a major web browser. It's even been growing steadily in popularity for years now, and has almost caught up with Firefox in market share.
You might be remembering when they migrated from their own engine to Chromium. The name Opera lives on, but as a contributing part of a diverse browser ecosystem it did die that day, as far as I’m...
You might be remembering when they migrated from their own engine to Chromium. The name Opera lives on, but as a contributing part of a diverse browser ecosystem it did die that day, as far as I’m concerned.
Opera didn’t migrate anything. They sold the browser and name to a Chinese firm who are now operating under that name, and they replaced Opera with their Chromium based product. The company...
Opera didn’t migrate anything. They sold the browser and name to a Chinese firm who are now operating under that name, and they replaced Opera with their Chromium based product. The company previously known as Opera is Otello.
I've mentioned this before, but Opera replaced their rendering engine (Presto) before they were acquired. Opera 15 with Chromium was released in 2013, while they were acquired in 2016.
I mean, Opera isn't Norwegian anymore. Hasn't been for many years. Vivaldi is a thing, based in Oslo and founded by one of the original founders of Opera.
I mean, Opera isn't Norwegian anymore. Hasn't been for many years.
Vivaldi is a thing, based in Oslo and founded by one of the original founders of Opera.
The modern day incarnation of Opera seems a bit sketchy to me, though I'm generally skeptical of any "free" software that has a real marketing budget. They are paying for that marketing somehow. A...
The modern day incarnation of Opera seems a bit sketchy to me, though I'm generally skeptical of any "free" software that has a real marketing budget. They are paying for that marketing somehow.
A number of OG Opera guys including a co-founder and former CEO are now running Vivaldi.
Maybe, some browsers even end up costing money/requiring some other form of licensing, even if it’s “only” for corporate settings or whatever – thus achieving even a hope user interests will...
Maybe, some browsers even end up costing money/requiring some other form of licensing, even if it’s “only” for corporate settings or whatever – thus achieving even a hope user interests will receive priority again.
The economics for developing another browser engine don't pencil out. There are many web APIs to implement in order for an alternative to be minimally functional in displaying modern web content....
The economics for developing another browser engine don't pencil out. There are many web APIs to implement in order for an alternative to be minimally functional in displaying modern web content. It's a long, expensive, difficult slog; and in the end, if you adhere faithfully to Web standards, you end up with an engine that'll be functionally the same as the incumbent engines.
However... if a viable alternative does emerge, I'd bet $100 that it'll be a state-directed project from China, which are politically motivated to control as much of its information environment as possible and on the basis of security want to de-risk relying on US-made browser engines. And if you're privacy-concerned, you're definitely not going to want to use theirs.
oof Dang. That sounds plausible. I had never considered how China factors into all of this. They seem to be determined to become fully independent. It makes me wonder what browsers people use there. 🤔
However... if a viable alternative does emerge, I'd bet $100 that it'll be a state-directed project from China, which are politically motivated to control as much of its information environment as possible and on the basis of security want to de-risk relying on US-made browser engines. And if you're privacy-concerned, you're definitely not going to want to use theirs.
oof Dang. That sounds plausible. I had never considered how China factors into all of this. They seem to be determined to become fully independent. It makes me wonder what browsers people use there. 🤔
Mostly the same as the rest of the world, but they do have a few local Chromium-based browsers (UC, 360 Safe, QQ) that together make up ~15% market share.
It makes me wonder what browsers people use there.
That's true, but https://ladybird.org/ is kind of proving that you don't need to support 100% of everything to have something usable. So my hope is that if a project like this can exist, maybe...
The economics for developing another browser engine don't pencil out. There are many web APIs to implement in order for an alternative to be minimally functional in displaying modern web content.
That's true, but https://ladybird.org/ is kind of proving that you don't need to support 100% of everything to have something usable. So my hope is that if a project like this can exist, maybe Mozilla also can? With user support instead of the mistake they made of relying on Google so much and all other bad decisions their leadership make.
Ladybird's first commit was in October 2018, and they don't expect to have their first alpha release until the summer of 2026, nearly 8 years into development. I think that shows just how...
Ladybird's first commit was in October 2018, and they don't expect to have their first alpha release until the summer of 2026, nearly 8 years into development. I think that shows just how difficult it is to build a browser engine from scratch. Firefox already has a complete engine, but that's still a lot to maintain and update without anyone paid to work on it full time.
For several years, Ladybird was a project within the larger SerenityOS project, and not Kling's main focus/its own thing. I think that makes the date of its first commit much less meaningful....
For several years, Ladybird was a project within the larger SerenityOS project, and not Kling's main focus/its own thing. I think that makes the date of its first commit much less meaningful.
The project has a newsletter that tracks Ladybird's conformance and completion against other browsers/engines in various test suites, if you want to keep up with its progress yourself.
I don't know if that's so bad. It still spreads out maintenance and decision-making. Is Linux functionally the same between distros? Depends who you ask. Not that that ecosystem is without issues
and in the end, if you adhere faithfully to Web standards, you end up with an engine that'll be functionally the same as the incumbent engines.
I don't know if that's so bad. It still spreads out maintenance and decision-making. Is Linux functionally the same between distros? Depends who you ask. Not that that ecosystem is without issues
Bad if you want this new competitor to keep being supported. We're well past the time where something like Linux could pop up and be a legit step competitor and gather a massive group of talented...
I don't know if that's so bad.
Bad if you want this new competitor to keep being supported. We're well past the time where something like Linux could pop up and be a legit step competitor and gather a massive group of talented maintainers. People who want to do that get poached for 200l+/yr at Google Apple Mozilla.
Is Linux functionally the same between distro
There's three layers to this: the Linux core kernel, the disfro forks (Debian, Unity, etc.) and then any OS built on top of that fork.
To reflect that in OS'S: Since we don't have a "core" in web, all the distros (servo, chromium, gecko) are cores in themselves. Making a new one of these is very difficult.
I'm wondering if AI will be able to step in and make alternative browser development more viable, at some point in the near-ish future. I've been using the agentic Claude Code CLI and it's very...
I'm wondering if AI will be able to step in and make alternative browser development more viable, at some point in the near-ish future. I've been using the agentic Claude Code CLI and it's very impressive what's possible with it. The real power is its ability to queue up work for itself, find information in a codebase, and iterate toward a solution when something doesn't work. This allows it to remain goal-oriented and self-correct as needed.
It's not perfect yet, but ask me again in five years. Heck, ask me in 6 months. I can absolutely see something like this forking an existing open-source browser, porting the codebase to an entirely different tech stack, then crawling through file-by-file to optimize and refactor to such a degree that it's no longer recognizable as the original program, yet has an equally capable feature set. If the originating codebase contains high test coverage with well-written tests (as I expect is necessary for any current-gen browser, due to the complexity of the software) then those can provide a really good framework for the AI to follow and verify its work afterward.
At the rate this tech is maturing I honestly think this could be possible in 5 years, and that feels like a conservative estimate to me.
Unless it's open source, just as they did with their Deepseek model. And I would really love and appreciate that.
However... if a viable alternative does emerge, I'd bet $100 that it'll be a state-directed project from China, which are politically motivated to control as much of its information environment as possible and on the basis of security want to de-risk relying on US-made browser engines. And if you're privacy-concerned, you're definitely not going to want to use theirs.
Unless it's open source, just as they did with their Deepseek model. And I would really love and appreciate that.
Unfortunately the head dev of ladybird shot himself in the foot. He rejected a PR that would update documentation with gender-neutral pronouns, which stirred up controversy.
Instead of relying on some random, poorly sourced blog post, here's the original pull request: https://github.com/SerenityOS/serenity/pull/6814 That pull request was opened on May 2nd, 2021. Kling...
Instead of relying on some random, poorly sourced blog post, here's the original pull request:
That pull request was opened on May 2nd, 2021. Kling responded and closed it the same day.
Over three years later, on July 2nd, 2024, that pull request was posted on Mastodon. That's when the "controversy" began. As far as I know, Kling's only involvement or communication throughout it all consisted exclusively of his comment at the top of the pull request and his action in closing it.
I think it's also worth noting, for those who don't click through the link, that the rejected PR was updating language referring to a generic/unspecified individual to use they/them rather than...
I think it's also worth noting, for those who don't click through the link, that the rejected PR was updating language referring to a generic/unspecified individual to use they/them rather than he/him -- a circumstance in which singular-they is an incredibly longstanding and commonly-used gender-neutral choice and in which it's the the politically-neutral polite norm in almost every English-speaking context (including the style guides of most major news organizations these days). Using he/him in generic contexts as the documentation did (and presumably still does) is by far the more politically-charged choice and refusing to change it further signals a refusal to even do the bare minimum to not exclude women and non-binary individuals.
That graph doesn't have any data about singular they, only about the occurrence "he or she", which is a much more recent invention. Naturally, singular they cannot be identified in the simple word...
That graph doesn't have any data about singular they, only about the occurrence "he or she", which is a much more recent invention. Naturally, singular they cannot be identified in the simple word search in Google books like that, but moreover, it predates the earliest books included in that dataset by a considerable margin. Singular they is only remotely recent when used to refer to a specific, known person (as in the case of my preferred pronouns). Its use for a generic person or someone of unknown gender, as in the documentation in question, is found in Chaucer and is older than the use of singular "you." It was the standard construction for that concept throughout English both then and thereafter. The introduction of "he or she" or simply "he" as an alternative occurred much later, and the insistence that singular they was "wrong" during the 19th and early 20th centuries was an ahistorical insistence on applying rules that have never been fundamentally part of English grammar in order to better conform to the grammar of Latin, a completely different language. Even during this time, examples of singilar they abound in English literature and it was used more or less without a care in speech.
Moreover, even if singular they were so recent that it did not become accepted at all until 10-15 years ago, Ladybird is not old enough to use that as its justification. It had been the norm for years by the time Ladybird even began. The project was begun well after major news organizations had adopted it in their style guides and it was widely used as the standard gender-neutral alternative before Ladybird existed. Refusing to change from exclusively "he" pronouns (they were not using even the more gender-neutral but still antiquated "he or she") is absolutely taking a strong stance in the 2020s. Even if the creators were exclusively non-native English speakers who had been taught (incorrectly) not to use singular they in school, which does still happen, refusing to back down and adding clauses about "political neutrality" to their ToS make it clear that they considered the bare minimum to be a hard line here.
In addition, the refusal to accept a PR for what amounts to changing a small grammatical error in the documentation in terms of its scope indicates to me that they're incredibly unwilling to accept even the smallest criticism. Even if there were no gendered element here and I had no reason to suspect the maintainers of Ladybird were misogynists or anti-woke chuds, refusing to back down or accept any criticism on such a minor change that affects the understandability of their documentation in no negative way is something that would cause me to more or less lose all faith in them as software engineers. If this is how they react to something so minor, I can't imagine how poorly they handle constructive criticism of their actual code.
Having dealt with this going over rules for Twilight Imperium, it just... drives me insane. It makes me (a dude) feel weird while reading it around the table. What is the point of this kind of...
Having dealt with this going over rules for Twilight Imperium, it just... drives me insane. It makes me (a dude) feel weird while reading it around the table. What is the point of this kind of alienation? I know why they do it, but why do they feel compelled to manifest this kind of thing?
Sorry this was a bit of an off topic rant, I'm just tired of people with technical knowledge being shitty human beings.
It's an issue, but it's a minor issue. I feel like discussion on these issues very often details entire projects and that's ultimately a really huge loss. I agree, language should be as inclusive...
Exemplary
It's an issue, but it's a minor issue. I feel like discussion on these issues very often details entire projects and that's ultimately a really huge loss.
I agree, language should be as inclusive as is practically possible, and I disagree with decisions made that go against it, but it's a mild disagree. I feel like that sort of subtlety really often gets lost in online discourse about social topics. There are varying degrees of bad, between someone was very slightly rude one time all the way up to actual genocide.
Stuff like pronouns not being as inclusive as they could be is far on the "less bad" side of the house. The best course of action is to voice your disagreement, then move onto more important things. We'd all be better off if we stopped giving minor problems like this so much energy and attention.
The fact that they refused to concede and took such a hard stance against any change on such an ultimately minor issue is a major issue when it comes to my faith in the leadership of the project,...
It's an issue, but it's a minor issue.
The fact that they refused to concede and took such a hard stance against any change on such an ultimately minor issue is a major issue when it comes to my faith in the leadership of the project, though. Why should I trust someone who refuses to change things or listen to criticism to write code? I don't want to use a web browser written by someone who is so unwilling to change even a minor thing -- who knows what that looks like in code review!
Makes it really hard to enjoy The IT Crowd and Harry Potter, too. 😢 I just want to broadcast a PSA to everybody who makes stuff for the public — every writer, artist, musician, actor, director,...
Makes it really hard to enjoy The IT Crowd and Harry Potter, too. 😢
I just want to broadcast a PSA to everybody who makes stuff for the public — every writer, artist, musician, actor, director, producer, game designer, programmer, etc. — just shut up and make great stuff. Resist whatever urge you have to flap your jaw in a way that denigrates any group of human beings. Doesn't matter how justified you feel in doing so. Don't ruin your legacy by alienating swathes of your customer base with abrasive grandstanding.
What's the controversy there? That's how I felt when I learned that the drummer from Cage the Elephant is an anti-vaxxer. They were probably my 2nd favorite band but then I found a bunch of rants...
The IT Crowd
What's the controversy there?
Resist whatever urge you have to flap your jaw in a way that denigrates any group of human beings
That's how I felt when I learned that the drummer from Cage the Elephant is an anti-vaxxer. They were probably my 2nd favorite band but then I found a bunch of rants on Reddit from him about how the Covid vaccine is fake and they'd never require vaccines or masks at their shows (he was saying this in 2021... so still very much in the midst of Covid).
A quick Wikipedia check suggests that the writer and director of The IT Crowd, Graham Linehan is described in the first sentence of his page as an "anti-transgender activist." And yeah, looking...
A quick Wikipedia check suggests that the writer and director of The IT Crowd,Graham Linehan is described in the first sentence of his page as an "anti-transgender activist." And yeah, looking into the corresponding section on the bio...very much so.
I built it on Mac totally fine this morning (commit 60bd5012). Didn't really try too much, though, so I don't have anything to say about how it is to actually use.
I built it on Mac totally fine this morning (commit 60bd5012). Didn't really try too much, though, so I don't have anything to say about how it is to actually use.
I’m absolutely supportive of adding new engines into the mix, but I think that “just” pushing engine market share towards a more even 33/33/33 type split would constitute a dramatic improvement....
I’m absolutely supportive of adding new engines into the mix, but I think that “just” pushing engine market share towards a more even 33/33/33 type split would constitute a dramatic improvement. Gecko is of course totally unrelated to the other two, but Blink has also diverged from WebKit to a very high degree and so WebKit serves as competition almost as well as Gecko does, especially when considering Apple’s differing priorities. For example Gecko and Blink don’t put much focus on battery life where that’s one of the top considerations in WebKit.
Maybe this is complacency or lack of imagination, but I disagree. Web browsers seem pretty mature and it’s unclear to me that we’d be missing much if browsers stopped changing. Standard web API’s...
Maybe this is complacency or lack of imagination, but I disagree. Web browsers seem pretty mature and it’s unclear to me that we’d be missing much if browsers stopped changing. Standard web API’s are pretty good. For most web sites, new web API’s can be ignored. All we really need are security updates.
Users and most developers don’t care which browser technology you use. If every browser used v8 for JavaScript (for example), it would be fine. It rarely matters, but testing would get a bit easier if they were all the same.
On the other hand, many websites are terrible. But it’s not the browser’s fault that they are so bloated. It’s not technical limitations that make them terrible, it’s market incentives to put video ads everywhere.
What this suggests to me is that is if all browsers were built from a common, open source code base that didn’t change much, it would probably be fine. It would be like Linux.
I 100% agree. Almost every use case that an operating system can do has been katamaried into a browser. At this point, a browser is more akin to a windowing system for Docker containers than a...
I 100% agree. Almost every use case that an operating system can do has been katamaried into a browser. At this point, a browser is more akin to a windowing system for Docker containers than a text rendering engine.
If an 80% funding cut were to be the thing that caused the perpetual churn of functionality to end, I'd see it as a good thing. It would give time for other competing engines to be rewritten from the ground up. It would allow developers to stick with a non-moving target of functionality.
In reality, there will probably be something like the Chromium Foundation alongside Mozilla, and all of the big players will donate X amount to both in order to keep the antitrust at bay while not crippling the web.
There are differences that make some sites, some important like quickbooks, break in safari. i assume this is a standards issue, but having one browser so dominant makes it easy to be lazy.
There are differences that make some sites, some important like quickbooks, break in safari. i assume this is a standards issue, but having one browser so dominant makes it easy to be lazy.
This has always been the case. And one dominant browser also allows that browser to implement stuff without a standard and everything further spirals out of control. Peak web was when IE, Firefox,...
This has always been the case. And one dominant browser also allows that browser to implement stuff without a standard and everything further spirals out of control.
Peak web was when IE, Firefox, and Chrome all had a significant enough marketshare that most needed to fix breakage across all of them.
I was working in webdev in this era and for the most part if you developed for "anything but IE" your results were fairly solid in everything but IE as most features in most browsers followed...
Peak web was when IE, Firefox, and Chrome all had a significant enough marketshare that most needed to fix breakage across all of them.
I was working in webdev in this era and for the most part if you developed for "anything but IE" your results were fairly solid in everything but IE as most features in most browsers followed standards and carefully documented where they had "extended" past the standards.
Then you had to build a load of quirks/bodge files to fix each version of IE that was still in use, which was always painful and multiplied for more versions you tried to support. There was always a tendency for management to fight against this approach but in my experience it was always significantly less work in the long run.
The lazy approach, that a lot of organisation took, was to develop IE-first as they had majority market share and then usually not bother to complete the compatibility work for the other browsers afterwards. Beyond frantic patching when they inevitably found (after launch) a bunch of their stakeholders or power users were on Macs.
I'm not directly involved in frontend much these days, but I'm involved enough to know there is surprisingly little in the way of cross-browser bodges needed these days for the vast majority of cases.
Yeah these days most actual incompatibilities are surfaced by things like doing weird stuff with iframes or using more niche APIs that only a handful of sites have any reason to use (and thus,...
Yeah these days most actual incompatibilities are surfaced by things like doing weird stuff with iframes or using more niche APIs that only a handful of sites have any reason to use (and thus, haven’t received as much attention as the stuff that everybody uses). Back in the IE days it was more things like radically different layout behavior.
Usually if a site is partially or wholly broken in non-Blink browsers it’s because critical functions of the site hinge on some subtle behavior difference that’s specific to Blink, which occurs because devs are now doing the thing with Chrome that they used to do with IE where they don’t test against anything else unless they’re absolutely forced to (e.g. the target audience disproportionately uses Apple products and thus good Safari support is non-negotiable).
Yes, there are occasionally problems like that due to lack of browser compatibility testing. On the other hand, if there were only one open source implementation, like is true of Electron, maybe...
Yes, there are occasionally problems like that due to lack of browser compatibility testing. On the other hand, if there were only one open source implementation, like is true of Electron, maybe such bugs would be even less common and laziness would be okay?
Websites need to conform to web standards because there are multiple browsers. We need multiple browsers with significant market share to ensure web sites conform to standards. It seems a bit circular?
I’m not sure how serious I am about that argument, but It seems plausible enough that I’m not entirely convinced that the usual arguments for web standards are still important.
Problem is that the closest thing to that would be Chromium. Some well funded fork would become more popular and people would design around the most full featured version.
Problem is that the closest thing to that would be Chromium. Some well funded fork would become more popular and people would design around the most full featured version.
Distribution is a big issue because security is as important for browsers as it is for OSes. We don’t want people downloading browsers from random untrusted vendors, and we do want people to trust...
Distribution is a big issue because security is as important for browsers as it is for OSes. We don’t want people downloading browsers from random untrusted vendors, and we do want people to trust security updates. This is high stakes. Browsers get attacked from all sides, from well-funded attackers.
There are many people who distrust Google, Apple, Microsoft, and Mozilla, for various reasons. But they do have competent security teams and infrastructure. It could be a lot worse.
How could it be improved? What organization(s) could become both more trusted and worthy of that trust?
Reproducible builds might help, but it’s only a small part of the problem.
This is such a ridiculously simplistic take. Apple and Microsoft are both capable of supporting ongoing development of their browsers without Google's bribe to set the default search engine. The biggest impact is on Mozilla, which doesn't have other income streams of any similar size.
There's going to be big impacts on the ecosystem, but it's going to be an effect of unbundling a 2 decade long campaign for Google to dominate online search and advertising.
Ok, I'm glad I'm not the only one with this initial reaction. This is well outside my interests and areas of knowledge so I was really confused by the framing of this article. Apple and Microsoft...
Ok, I'm glad I'm not the only one with this initial reaction. This is well outside my interests and areas of knowledge so I was really confused by the framing of this article. Apple and Microsoft are multi-trillion dollar companies, I think they'll be ok.
I still have a ton of other questions and am looking for a good breakdown for dumbasses like me, preferably one with a corkboard and some red string.
Ugh, the Google funding has been a sword of Damocles for Mozilla for as long as I can remember. It was only a matter of time before it dropped. I always assumed Google would wield it as leverage...
Ugh, the Google funding has been a sword of Damocles for Mozilla for as long as I can remember. It was only a matter of time before it dropped. I always assumed Google would wield it as leverage for themselves, not that they would be compelled to take it away, but the end result is about the same for Firefox. I’ll be immeasurably disappointed if Mozilla didn’t prepare for this eventuality, somehow.
It won't happen but maybe the alternative path taken could have been simplifying web standards to a point where a web browser does not have to be effectively as complicated as an os. Suddenly the...
It won't happen but maybe the alternative path taken could have been simplifying web standards to a point where a web browser does not have to be effectively as complicated as an os. Suddenly the cost for developing a browser drops by that eighty percent easily and no such questionable income streams are necessary.
I think there actually is some niche project for either a small subset of sites or a niche protocol that aims to be a minimal way a view a web but I can't remember the name.
I think there actually is some niche project for either a small subset of sites or a niche protocol that aims to be a minimal way a view a web but I can't remember the name.
I think you're thinking of Gemini. I agreed with its goals, but I never got into it because it's... extremely opinionated, to the point of forbidding a lot of reasonable content. For example, it...
I agreed with its goals, but I never got into it because it's... extremely opinionated, to the point of forbidding a lot of reasonable content. For example, it doesn't support inline links. I couldn't have written that first line in a Gemini page. I'd have to instead do:
The 80% loss of funding assumes no other search engines want to bid for the position of default search engine for any of these browsers, which is frankly ridiculous. Revenue for web browsers will...
The 80% loss of funding assumes no other search engines want to bid for the position of default search engine for any of these browsers, which is frankly ridiculous. Revenue for web browsers will drop with one less search engine in the bidding, but not by nearly that much.
Well as you acknowledged since Google almost surely won't be allowed to bid on the default search engine, that's one huge competitor out of the way for remaining search engines. Beyond this, isn't...
Well as you acknowledged since Google almost surely won't be allowed to bid on the default search engine, that's one huge competitor out of the way for remaining search engines. Beyond this, isn't part of the governments case against Google that they were using their dominant position on search and their advertising side to jack up the cost of advertising and then using those extra funds to overpay to maintain their semi-monopoly? Basically all of the companies developing these browsers were financially benefitting from Googles ad dominance through search.
That's my understanding of the reason other search engines couldn't compete with Google on bids to be the default, because they didn't have leverage the way Google did to jack up costs for advertising.
So couple that together with Google not even being able to throw their hat into the ring, I could very well believe it's a substantial reduction in the amount of money they can get in new search deals.
I will say perhaps the biggest change coming up is AI results and perhaps that is what will drive up bids, considering AI is threatening to upend the whole web search model as it is, this could be the inflection point to shift the power of what collectively people default to when looking for answers or information.
The main thing I question about this is that I'm not sure if any of them yet have a solid revenue model to justify the large bids. That doesn't mean some of them backed with all their venture capital or billionaire backers won't try to throw that money at it, but I question the viableness of that funding in the long term for browsers if they can't actually find a way to generate the revenue to continue justifying those deals.
I'm not sure the benefits will outweigh the pain, but I'm also not certain that the pain will be excessive. It seems to me that browsers are a relatively mature product. I'm struggling to think of...
I'm not sure the benefits will outweigh the pain, but I'm also not certain that the pain will be excessive. It seems to me that browsers are a relatively mature product. I'm struggling to think of major innovations in recent years.
I remember the early days when sites would target the functionality in different browsers. It seems for quite a while now that for the most part, nobody worries about what browser someone uses except in somewhat edge cases. If nothing else, many frameworks tend to account for the differences and doing a "css reset" was a thing - it's not something I think about anymore (but I use two different frameworks).
My main concern these days is dealing with making a site responsive at various screen sizes and ratios. But even that has matured at this point.
I'm not saying someone can't come up with things to innovate with, but I think it's more incremental these days.
On the side of the browser design as software - again, I think the innovations are slower and less huge in nature. I liked how Edge brought solidly fantastic vertical tabs into the mix. Firefox now has native tab support that works for me - great timing as I've started having troubles with Edge being unstable. So I'm back on Firefox.
It was an easy switch - import my bookmarks and saved passwords and I'm pretty much good to go with a bit of tweaking. The most critical extension for me is uBlock Origin, so I moved from a browser that supports it to another browser that supports it.
Do we really need more page rendering engines? It takes a huge amount of effort to develop those. I'm not sure there's much compelling stuff left to do. Sure, maybe someone will come up with an awesome idea. If we can continue to get bugfixes, that will just make the web slightly better and better.
And sure, like vertical tabs coming to Firefox, maybe that's mini-innovation?
I don't know. I realize such development may slow down, but I think we're at a point that for the most part, that should be alright.
I think the only really big innovations will be somehow surrounding AI, and I think most people are not so eager for those anyway. Most of the AI stuff has been quite annoying — and I say that as someone who uses AI regularly for some tasks. It's great for what it's great for, but it's generally annoying to see it pop up everywhere. Even if I begrudgingly admit that a lot of what it's doing out there is actually useful. And as long as it continues to improve — well, that's another topic entirely.
You’re describing user-facing features, but it’s not like the web platform has peaked. The various JS APIs and CSS spec are constantly evolving, with no sign of slowing down. Supporting that is...
You’re describing user-facing features, but it’s not like the web platform has peaked. The various JS APIs and CSS spec are constantly evolving, with no sign of slowing down. Supporting that is the bulk of modern browser development. Those things are invisible to most users but (along with security patches) are vital to the health of the ecosystem.
Agreed. CSS in particular has received some major improvements in the last two years alone. Nesting rules, the :has() selector, container queries, and @scope are groundbreaking. CSS Color Module...
Agreed. CSS in particular has received some major improvements in the last two years alone. Nesting rules, the :has() selector, container queries, and @scope are groundbreaking. CSS Color Module Level 4 has also brought us new colour spaces, wider gamut support (hello HDR), and the incredible color-mix().
Browsers are under constant development. Users might not see it, but developers certainly do.
There are people who build websites but aren’t full time web developers and aren’t even aware of any of that. We barely know CSS. Tailwind is popular because it makes knowing CSS someone else’s...
There are people who build websites but aren’t full time web developers and aren’t even aware of any of that. We barely know CSS. Tailwind is popular because it makes knowing CSS someone else’s problem.
We certainly won’t care about color spaces. HDR? Yawn. The web was already colorful enough.
I appreciate the point, but I can't say I agree, Brian. Tailwind provides a lot of value, but it's not because it saves you from learning CSS. You still need to understand layout models (flow,...
I appreciate the point, but I can't say I agree, Brian. Tailwind provides a lot of value, but it's not because it saves you from learning CSS. You still need to understand layout models (flow, float, flex, grid, et al). You still deal with inheritance via the cascade. You still need to understand specificity. TW isn't a deep abstraction over CSS.
I would argue that Tailwind is primarily popular because it's exceptionally well-suited to component-based design, provides standardization among teams, and is designed for build pipelines with features like tree shaking. It's the evolution of earlier approaches like BEM to writing tightly-scoped components without unexpected side effects everywhere. Atomic classes just integrate better into modern JS stacks like React and Vue.
I don't disagree that there are tons of people building pages using whatever tools they have available. Many of them are creating sites for their small business, or just making a homepage for their dog. Though I suspect most of them are using apps like Webflow and Wix, rather than CSS frameworks like Tailwind or even Bootstrap. So while they may not use these new CSS features directly, the tools like Wix they rely on are using them under the hood.
I think what we're really discussing here though is if these kinds of features should be moved from the page-level to the browser-level. A lot of features or complex layouts were possible in the past, but they might have required JS shims, workarounds, or other hacks. Browser vendors typically look at what people are trying to accomplish but can't, or what they do implement but in a clunky or poor way. Then they review solutions to make that better.
While I get the concern about making browsers too complex, I also think it makes sense to put the complexity in the browser, and not leave it up to every single page to implement in their own, non-standard way. Browsers are constantly evolving to optimize earlier functions, and even new functionality and enhancements can be added along the way. Some examples are alert() windows adding an "ignore future alerts" checkbox, text areas becoming resizable, and form submissions adding retry prompts. Accessibility also plays a large role, as new features like <dialog> can be properly contextualized by screen readers rather than trying to guess at their meaning.
In my opinion, websites are getting larger, but they should be getting smaller as popular features require less code to implement. Today I can create an effect where clicking a search button smoothly expands an input bar to fill the available space in under ten lines of CSS. Something like that would have required far more expensive JS and layout logic in the past. Websites being bloated and slow to load comes largely from wasted code, over-engineering, and poor priorities during development. It's not because browsers are too capable.
That said, I do think some bad APIs have made it into the browser. Web components are a bit faff, and IndexedDB was a bust. Arguably, the entire cascade model is not very good. But most APIs are well-researched and agreed upon via consensus, and are targeted to reduce specific developer pain points. All of the additions I described above were great candidates.
So when is it enough? That's a fair question, but realistically I don't think the web will ever be "done". It constantly evolves to meet new needs. Before 2007, nobody browsed the web by panning on touch screens. Yet overnight, the entire idea of a "cursor" disappeared, and we needed to rethink layout, scaling, and interaction. We addressed that with features like media queries, image srcset, and multi-touch, but the effort is still ongoing.
Who can predict what's next? What does the web look like in AR or VR, where we have gestures but no keyboard? When navigating through voice via wireless earbuds? By engaging with an AI agent? Or browsed via brain-computer interface? Interesting questions, and some we may need to answer eventually.
I think we can agree that the web has never been static. From the earliest support for tables and images, it has grown to meet our needs. As long as people continue to want to create and share experience online, I expect that will continue. Whether that means building something complex, or just making a homepage for our dog.
I do think the web standards folks are doing good work, and there are nice improvements. I'm not against that work continuing. I'm just saying, stagnation wouldn't be so bad. If you had to build...
I do think the web standards folks are doing good work, and there are nice improvements. I'm not against that work continuing. I'm just saying, stagnation wouldn't be so bad.
If you had to build an ordinary website using only API's available five years ago, it seems like it wouldn't be too painful?
Five years ago wasn't so bad. Ten years definitely had pain points. Twenty years ago, oh boy. So I can take your point that all the "major issues" have been worked out, and slowing the pace might...
Five years ago wasn't so bad. Ten years definitely had pain points. Twenty years ago, oh boy.
So I can take your point that all the "major issues" have been worked out, and slowing the pace might be okay. But I also think it's the kind of assessment we can only really make in the context of the present. Twenty years ago, I thought it wasn't so bad using spacer.gif images to control table layouts. Twenty years from now, will I equally look back on today as the "dark ages"?
For example, I wouldn't be surprised if we consider untyped JS to be reckless, cowboy programming. Or scoff at the amount of boilerplate required to add "simple things" like data bindings, in the same way we talk about jQuery now. But it's really hard to say without hindsight.
I'd also mention that it's not just about big changes. A lot of the recent focus has been on serving smaller groups, with features like enhanced accessibility controls, right-to-left formatting, better support for uncommon languages, and sensitivity toggles like prefers-reduced-motion. With the web being such a fundamental aspect of our modern lives, it needs to be usable by everyone. A lot of work goes into making that happen.
Realistically, I think stagnation only works for so long. If we collectively took a 5 year breather, then no, I don't think anything really bad would happen. But the problems would start to pile up. For users, features like passkeys would languish, because why use something that doesn't integrate well into your digital life? For developers, tools like style queries would only be partially implemented, and would be considered buggy/unreliable. And while you dismissed HDR colors before, apps like Pixlr need that to be able to work accurately with your new digital cameras.
I don't think stagnation kills the web, exactly, but it would certainly be less lively than before.
Yep, I agree. Also, there are likely lots of other improvements important to certain people or essential for certain specialized apps (like Pixlr) that would be unnoticed by most.
Yep, I agree. Also, there are likely lots of other improvements important to certain people or essential for certain specialized apps (like Pixlr) that would be unnoticed by most.
Agreed. See this problem and this proposed solution as an example. This would make building a small, personal site so much easier for amateurs and people who don't want to use JavaScript.
Agreed. See this problem and this proposed solution as an example. This would make building a small, personal site so much easier for amateurs and people who don't want to use JavaScript.
If web browsers are invested in less and fall behind, then companies will invest more in the closed ecosystems of iOS and Android apps instead. I don't think people appreciate the web ecosystem...
If web browsers are invested in less and fall behind, then companies will invest more in the closed ecosystems of iOS and Android apps instead. I don't think people appreciate the web ecosystem enough for being an open standards-based system with multiple interoperable implementations. Twenty years ago the ecosystem was bad and it was a crazy fantasy that it could ever get as good as it is now; I hope it doesn't peak and taper off here now.
This seems kinda like concern trolling? I agree what more competition in the web browser space. But this feels like "oh y'all were so gung ho to go after Google's anti competitive payments to...
This seems kinda like concern trolling? I agree what more competition in the web browser space. But this feels like "oh y'all were so gung ho to go after Google's anti competitive payments to protect their search engine. Now you're gonna get you comeuppance!"
Sounds like the big browsers will be competing with the smaller competitors who didn't have those kickbacks?
Can we really call edge a major browser in this context when its another chromium based browser? I'm sure there's more to it than slapping an edge mask on chromium but still.
As a Norwegian, I'm a bit sad that Opera is no longer considered a major web browser. It's even been growing steadily in popularity for years now, and has almost caught up with Firefox in market share.
I honestly thought Opera got shut down like 10 years ago
You might be remembering when they migrated from their own engine to Chromium. The name Opera lives on, but as a contributing part of a diverse browser ecosystem it did die that day, as far as I’m concerned.
Vivaldi is the real successor of opera, made by the same guys. But it's also chromium based.
Opera didn’t migrate anything. They sold the browser and name to a Chinese firm who are now operating under that name, and they replaced Opera with their Chromium based product. The company previously known as Opera is Otello.
I've mentioned this before, but Opera replaced their rendering engine (Presto) before they were acquired. Opera 15 with Chromium was released in 2013, while they were acquired in 2016.
Oh, I hadn’t heard of that. What a shame.
For sure. Now opera to me is the ad riddled browser all my friends use for some reason.
I mean, Opera isn't Norwegian anymore. Hasn't been for many years.
Vivaldi is a thing, based in Oslo and founded by one of the original founders of Opera.
Made with ❤️ in Europe!
The modern day incarnation of Opera seems a bit sketchy to me, though I'm generally skeptical of any "free" software that has a real marketing budget. They are paying for that marketing somehow.
A number of OG Opera guys including a co-founder and former CEO are now running Vivaldi.
I would love this. If it was even mildly functional, I think that I’d jump onto it with all my heart.
Maybe, some browsers even end up costing money/requiring some other form of licensing, even if it’s “only” for corporate settings or whatever – thus achieving even a hope user interests will receive priority again.
Boy oh boy do we need real alternatives to Chromium, Gecko, and Webkit.
The economics for developing another browser engine don't pencil out. There are many web APIs to implement in order for an alternative to be minimally functional in displaying modern web content. It's a long, expensive, difficult slog; and in the end, if you adhere faithfully to Web standards, you end up with an engine that'll be functionally the same as the incumbent engines.
However... if a viable alternative does emerge, I'd bet $100 that it'll be a state-directed project from China, which are politically motivated to control as much of its information environment as possible and on the basis of security want to de-risk relying on US-made browser engines. And if you're privacy-concerned, you're definitely not going to want to use theirs.
oof Dang. That sounds plausible. I had never considered how China factors into all of this. They seem to be determined to become fully independent. It makes me wonder what browsers people use there. 🤔
Mostly the same as the rest of the world, but they do have a few local Chromium-based browsers (UC, 360 Safe, QQ) that together make up ~15% market share.
Wow, and here I thought the CCP would throw a person into a forced labor camp if they were caught using any Google products.
That's true, but https://ladybird.org/ is kind of proving that you don't need to support 100% of everything to have something usable. So my hope is that if a project like this can exist, maybe Mozilla also can? With user support instead of the mistake they made of relying on Google so much and all other bad decisions their leadership make.
Ladybird's first commit was in October 2018, and they don't expect to have their first alpha release until the summer of 2026, nearly 8 years into development. I think that shows just how difficult it is to build a browser engine from scratch. Firefox already has a complete engine, but that's still a lot to maintain and update without anyone paid to work on it full time.
For several years, Ladybird was a project within the larger SerenityOS project, and not Kling's main focus/its own thing. I think that makes the date of its first commit much less meaningful.
Considering Ladybird didn't become its own project until June 2024, I think an alpha release in 2026 is actually quite impressive.
The project has a newsletter that tracks Ladybird's conformance and completion against other browsers/engines in various test suites, if you want to keep up with its progress yourself.
It was also made mostly by one guy as only a part of his larger project. A proper focused team should be able to do it way faster.
I don't know if that's so bad. It still spreads out maintenance and decision-making. Is Linux functionally the same between distros? Depends who you ask. Not that that ecosystem is without issues
Bad if you want this new competitor to keep being supported. We're well past the time where something like Linux could pop up and be a legit step competitor and gather a massive group of talented maintainers. People who want to do that get poached for 200l+/yr at Google Apple Mozilla.
There's three layers to this: the Linux core kernel, the disfro forks (Debian, Unity, etc.) and then any OS built on top of that fork.
To reflect that in OS'S: Since we don't have a "core" in web, all the distros (servo, chromium, gecko) are cores in themselves. Making a new one of these is very difficult.
I'm wondering if AI will be able to step in and make alternative browser development more viable, at some point in the near-ish future. I've been using the agentic Claude Code CLI and it's very impressive what's possible with it. The real power is its ability to queue up work for itself, find information in a codebase, and iterate toward a solution when something doesn't work. This allows it to remain goal-oriented and self-correct as needed.
It's not perfect yet, but ask me again in five years. Heck, ask me in 6 months. I can absolutely see something like this forking an existing open-source browser, porting the codebase to an entirely different tech stack, then crawling through file-by-file to optimize and refactor to such a degree that it's no longer recognizable as the original program, yet has an equally capable feature set. If the originating codebase contains high test coverage with well-written tests (as I expect is necessary for any current-gen browser, due to the complexity of the software) then those can provide a really good framework for the AI to follow and verify its work afterward.
At the rate this tech is maturing I honestly think this could be possible in 5 years, and that feels like a conservative estimate to me.
Unless it's open source, just as they did with their Deepseek model. And I would really love and appreciate that.
Unfortunately the head dev of ladybird shot himself in the foot. He rejected a PR that would update documentation with gender-neutral pronouns, which stirred up controversy.
Instead of relying on some random, poorly sourced blog post, here's the original pull request:
https://github.com/SerenityOS/serenity/pull/6814
That pull request was opened on May 2nd, 2021. Kling responded and closed it the same day.
Over three years later, on July 2nd, 2024, that pull request was posted on Mastodon. That's when the "controversy" began. As far as I know, Kling's only involvement or communication throughout it all consisted exclusively of his comment at the top of the pull request and his action in closing it.
Note that the contributing guidelines now include these rules on language:
The commit that added the rule on gender-neutral pronouns was added by Kling himself on July 29, 2024:
https://github.com/LadybirdBrowser/ladybird/commit/627dcb90bdd23ccfa2ae210d55b474ab4a844db0
I think it's also worth noting, for those who don't click through the link, that the rejected PR was updating language referring to a generic/unspecified individual to use they/them rather than he/him -- a circumstance in which singular-they is an incredibly longstanding and commonly-used gender-neutral choice and in which it's the the politically-neutral polite norm in almost every English-speaking context (including the style guides of most major news organizations these days). Using he/him in generic contexts as the documentation did (and presumably still does) is by far the more politically-charged choice and refusing to change it further signals a refusal to even do the bare minimum to not exclude women and non-binary individuals.
Singular "they" really only started gaining popularity from "he" or "he or she" like 10–15 years ago and is still far from ubiquitous: https://books.google.com/ngrams/graph?content=he+or+she&year_start=1800&year_end=2022&corpus=en&smoothing=3
I don't think it's fair to frame this as some countercultural thing.
That graph doesn't have any data about singular they, only about the occurrence "he or she", which is a much more recent invention. Naturally, singular they cannot be identified in the simple word search in Google books like that, but moreover, it predates the earliest books included in that dataset by a considerable margin. Singular they is only remotely recent when used to refer to a specific, known person (as in the case of my preferred pronouns). Its use for a generic person or someone of unknown gender, as in the documentation in question, is found in Chaucer and is older than the use of singular "you." It was the standard construction for that concept throughout English both then and thereafter. The introduction of "he or she" or simply "he" as an alternative occurred much later, and the insistence that singular they was "wrong" during the 19th and early 20th centuries was an ahistorical insistence on applying rules that have never been fundamentally part of English grammar in order to better conform to the grammar of Latin, a completely different language. Even during this time, examples of singilar they abound in English literature and it was used more or less without a care in speech.
Moreover, even if singular they were so recent that it did not become accepted at all until 10-15 years ago, Ladybird is not old enough to use that as its justification. It had been the norm for years by the time Ladybird even began. The project was begun well after major news organizations had adopted it in their style guides and it was widely used as the standard gender-neutral alternative before Ladybird existed. Refusing to change from exclusively "he" pronouns (they were not using even the more gender-neutral but still antiquated "he or she") is absolutely taking a strong stance in the 2020s. Even if the creators were exclusively non-native English speakers who had been taught (incorrectly) not to use singular they in school, which does still happen, refusing to back down and adding clauses about "political neutrality" to their ToS make it clear that they considered the bare minimum to be a hard line here.
In addition, the refusal to accept a PR for what amounts to changing a small grammatical error in the documentation in terms of its scope indicates to me that they're incredibly unwilling to accept even the smallest criticism. Even if there were no gendered element here and I had no reason to suspect the maintainers of Ladybird were misogynists or anti-woke chuds, refusing to back down or accept any criticism on such a minor change that affects the understandability of their documentation in no negative way is something that would cause me to more or less lose all faith in them as software engineers. If this is how they react to something so minor, I can't imagine how poorly they handle constructive criticism of their actual code.
Having dealt with this going over rules for Twilight Imperium, it just... drives me insane. It makes me (a dude) feel weird while reading it around the table. What is the point of this kind of alienation? I know why they do it, but why do they feel compelled to manifest this kind of thing?
Sorry this was a bit of an off topic rant, I'm just tired of people with technical knowledge being shitty human beings.
It's an issue, but it's a minor issue. I feel like discussion on these issues very often details entire projects and that's ultimately a really huge loss.
I agree, language should be as inclusive as is practically possible, and I disagree with decisions made that go against it, but it's a mild disagree. I feel like that sort of subtlety really often gets lost in online discourse about social topics. There are varying degrees of bad, between someone was very slightly rude one time all the way up to actual genocide.
Stuff like pronouns not being as inclusive as they could be is far on the "less bad" side of the house. The best course of action is to voice your disagreement, then move onto more important things. We'd all be better off if we stopped giving minor problems like this so much energy and attention.
The fact that they refused to concede and took such a hard stance against any change on such an ultimately minor issue is a major issue when it comes to my faith in the leadership of the project, though. Why should I trust someone who refuses to change things or listen to criticism to write code? I don't want to use a web browser written by someone who is so unwilling to change even a minor thing -- who knows what that looks like in code review!
Makes it really hard to enjoy The IT Crowd and Harry Potter, too. 😢
I just want to broadcast a PSA to everybody who makes stuff for the public — every writer, artist, musician, actor, director, producer, game designer, programmer, etc. — just shut up and make great stuff. Resist whatever urge you have to flap your jaw in a way that denigrates any group of human beings. Doesn't matter how justified you feel in doing so. Don't ruin your legacy by alienating swathes of your customer base with abrasive grandstanding.
What's the controversy there?
That's how I felt when I learned that the drummer from Cage the Elephant is an anti-vaxxer. They were probably my 2nd favorite band but then I found a bunch of rants on Reddit from him about how the Covid vaccine is fake and they'd never require vaccines or masks at their shows (he was saying this in 2021... so still very much in the midst of Covid).
https://en.wikipedia.org/wiki/Graham_Linehan#Anti-transgender_activism
Thanks. That sucks
A quick Wikipedia check suggests that the writer and director of The IT Crowd, Graham Linehan is described in the first sentence of his page as an "anti-transgender activist." And yeah, looking into the corresponding section on the bio...very much so.
Thanks :/
Damn, that's a shame. Reminds me of the Nim lang dude...
How is it going? I haven’t tried it personally, I’m skeptical that it could ever become competitive, but I’m interested to hear your impressions.
I built it on Mac totally fine this morning (commit
60bd5012). Didn't really try too much, though, so I don't have anything to say about how it is to actually use.I’m absolutely supportive of adding new engines into the mix, but I think that “just” pushing engine market share towards a more even 33/33/33 type split would constitute a dramatic improvement. Gecko is of course totally unrelated to the other two, but Blink has also diverged from WebKit to a very high degree and so WebKit serves as competition almost as well as Gecko does, especially when considering Apple’s differing priorities. For example Gecko and Blink don’t put much focus on battery life where that’s one of the top considerations in WebKit.
Maybe this is complacency or lack of imagination, but I disagree. Web browsers seem pretty mature and it’s unclear to me that we’d be missing much if browsers stopped changing. Standard web API’s are pretty good. For most web sites, new web API’s can be ignored. All we really need are security updates.
Users and most developers don’t care which browser technology you use. If every browser used v8 for JavaScript (for example), it would be fine. It rarely matters, but testing would get a bit easier if they were all the same.
On the other hand, many websites are terrible. But it’s not the browser’s fault that they are so bloated. It’s not technical limitations that make them terrible, it’s market incentives to put video ads everywhere.
What this suggests to me is that is if all browsers were built from a common, open source code base that didn’t change much, it would probably be fine. It would be like Linux.
I 100% agree. Almost every use case that an operating system can do has been katamaried into a browser. At this point, a browser is more akin to a windowing system for Docker containers than a text rendering engine.
If an 80% funding cut were to be the thing that caused the perpetual churn of functionality to end, I'd see it as a good thing. It would give time for other competing engines to be rewritten from the ground up. It would allow developers to stick with a non-moving target of functionality.
In reality, there will probably be something like the Chromium Foundation alongside Mozilla, and all of the big players will donate X amount to both in order to keep the antitrust at bay while not crippling the web.
There are differences that make some sites, some important like quickbooks, break in safari. i assume this is a standards issue, but having one browser so dominant makes it easy to be lazy.
This has always been the case. And one dominant browser also allows that browser to implement stuff without a standard and everything further spirals out of control.
Peak web was when IE, Firefox, and Chrome all had a significant enough marketshare that most needed to fix breakage across all of them.
I was working in webdev in this era and for the most part if you developed for "anything but IE" your results were fairly solid in everything but IE as most features in most browsers followed standards and carefully documented where they had "extended" past the standards.
Then you had to build a load of quirks/bodge files to fix each version of IE that was still in use, which was always painful and multiplied for more versions you tried to support. There was always a tendency for management to fight against this approach but in my experience it was always significantly less work in the long run.
The lazy approach, that a lot of organisation took, was to develop IE-first as they had majority market share and then usually not bother to complete the compatibility work for the other browsers afterwards. Beyond frantic patching when they inevitably found (after launch) a bunch of their stakeholders or power users were on Macs.
I'm not directly involved in frontend much these days, but I'm involved enough to know there is surprisingly little in the way of cross-browser bodges needed these days for the vast majority of cases.
Yeah these days most actual incompatibilities are surfaced by things like doing weird stuff with iframes or using more niche APIs that only a handful of sites have any reason to use (and thus, haven’t received as much attention as the stuff that everybody uses). Back in the IE days it was more things like radically different layout behavior.
Usually if a site is partially or wholly broken in non-Blink browsers it’s because critical functions of the site hinge on some subtle behavior difference that’s specific to Blink, which occurs because devs are now doing the thing with Chrome that they used to do with IE where they don’t test against anything else unless they’re absolutely forced to (e.g. the target audience disproportionately uses Apple products and thus good Safari support is non-negotiable).
In my experience, 99.9% of broken websites in Firefox are fixed by putting in a Chrome user agent string. Just like in 2008ish.
Yes, there are occasionally problems like that due to lack of browser compatibility testing. On the other hand, if there were only one open source implementation, like is true of Electron, maybe such bugs would be even less common and laziness would be okay?
Websites need to conform to web standards because there are multiple browsers. We need multiple browsers with significant market share to ensure web sites conform to standards. It seems a bit circular?
I’m not sure how serious I am about that argument, but It seems plausible enough that I’m not entirely convinced that the usual arguments for web standards are still important.
Problem is that the closest thing to that would be Chromium. Some well funded fork would become more popular and people would design around the most full featured version.
Distribution is a big issue because security is as important for browsers as it is for OSes. We don’t want people downloading browsers from random untrusted vendors, and we do want people to trust security updates. This is high stakes. Browsers get attacked from all sides, from well-funded attackers.
There are many people who distrust Google, Apple, Microsoft, and Mozilla, for various reasons. But they do have competent security teams and infrastructure. It could be a lot worse.
How could it be improved? What organization(s) could become both more trusted and worthy of that trust?
Reproducible builds might help, but it’s only a small part of the problem.
From the comments, encapsulates my thoughts:
Ok, I'm glad I'm not the only one with this initial reaction. This is well outside my interests and areas of knowledge so I was really confused by the framing of this article. Apple and Microsoft are multi-trillion dollar companies, I think they'll be ok.
I still have a ton of other questions and am looking for a good breakdown for dumbasses like me, preferably one with a corkboard and some red string.
Ugh, the Google funding has been a sword of Damocles for Mozilla for as long as I can remember. It was only a matter of time before it dropped. I always assumed Google would wield it as leverage for themselves, not that they would be compelled to take it away, but the end result is about the same for Firefox. I’ll be immeasurably disappointed if Mozilla didn’t prepare for this eventuality, somehow.
It won't happen but maybe the alternative path taken could have been simplifying web standards to a point where a web browser does not have to be effectively as complicated as an os. Suddenly the cost for developing a browser drops by that eighty percent easily and no such questionable income streams are necessary.
I think there actually is some niche project for either a small subset of sites or a niche protocol that aims to be a minimal way a view a web but I can't remember the name.
Gopher? ;-)
get out of here. archie all day long.
I think you're thinking of Gemini.
I agreed with its goals, but I never got into it because it's... extremely opinionated, to the point of forbidding a lot of reasonable content. For example, it doesn't support inline links. I couldn't have written that first line in a Gemini page. I'd have to instead do:
I wonder what could have been had the authors not been so preoccupied with reproducing Gopher's limitations.
The 80% loss of funding assumes no other search engines want to bid for the position of default search engine for any of these browsers, which is frankly ridiculous. Revenue for web browsers will drop with one less search engine in the bidding, but not by nearly that much.
Well as you acknowledged since Google almost surely won't be allowed to bid on the default search engine, that's one huge competitor out of the way for remaining search engines. Beyond this, isn't part of the governments case against Google that they were using their dominant position on search and their advertising side to jack up the cost of advertising and then using those extra funds to overpay to maintain their semi-monopoly? Basically all of the companies developing these browsers were financially benefitting from Googles ad dominance through search.
That's my understanding of the reason other search engines couldn't compete with Google on bids to be the default, because they didn't have leverage the way Google did to jack up costs for advertising.
So couple that together with Google not even being able to throw their hat into the ring, I could very well believe it's a substantial reduction in the amount of money they can get in new search deals.
I will say perhaps the biggest change coming up is AI results and perhaps that is what will drive up bids, considering AI is threatening to upend the whole web search model as it is, this could be the inflection point to shift the power of what collectively people default to when looking for answers or information.
The main thing I question about this is that I'm not sure if any of them yet have a solid revenue model to justify the large bids. That doesn't mean some of them backed with all their venture capital or billionaire backers won't try to throw that money at it, but I question the viableness of that funding in the long term for browsers if they can't actually find a way to generate the revenue to continue justifying those deals.
I'm not sure the benefits will outweigh the pain, but I'm also not certain that the pain will be excessive. It seems to me that browsers are a relatively mature product. I'm struggling to think of major innovations in recent years.
I remember the early days when sites would target the functionality in different browsers. It seems for quite a while now that for the most part, nobody worries about what browser someone uses except in somewhat edge cases. If nothing else, many frameworks tend to account for the differences and doing a "css reset" was a thing - it's not something I think about anymore (but I use two different frameworks).
My main concern these days is dealing with making a site responsive at various screen sizes and ratios. But even that has matured at this point.
I'm not saying someone can't come up with things to innovate with, but I think it's more incremental these days.
On the side of the browser design as software - again, I think the innovations are slower and less huge in nature. I liked how Edge brought solidly fantastic vertical tabs into the mix. Firefox now has native tab support that works for me - great timing as I've started having troubles with Edge being unstable. So I'm back on Firefox.
It was an easy switch - import my bookmarks and saved passwords and I'm pretty much good to go with a bit of tweaking. The most critical extension for me is uBlock Origin, so I moved from a browser that supports it to another browser that supports it.
Do we really need more page rendering engines? It takes a huge amount of effort to develop those. I'm not sure there's much compelling stuff left to do. Sure, maybe someone will come up with an awesome idea. If we can continue to get bugfixes, that will just make the web slightly better and better.
And sure, like vertical tabs coming to Firefox, maybe that's mini-innovation?
I don't know. I realize such development may slow down, but I think we're at a point that for the most part, that should be alright.
I think the only really big innovations will be somehow surrounding AI, and I think most people are not so eager for those anyway. Most of the AI stuff has been quite annoying — and I say that as someone who uses AI regularly for some tasks. It's great for what it's great for, but it's generally annoying to see it pop up everywhere. Even if I begrudgingly admit that a lot of what it's doing out there is actually useful. And as long as it continues to improve — well, that's another topic entirely.
You’re describing user-facing features, but it’s not like the web platform has peaked. The various JS APIs and CSS spec are constantly evolving, with no sign of slowing down. Supporting that is the bulk of modern browser development. Those things are invisible to most users but (along with security patches) are vital to the health of the ecosystem.
Agreed. CSS in particular has received some major improvements in the last two years alone. Nesting rules, the
:has()selector, container queries, and @scope are groundbreaking. CSS Color Module Level 4 has also brought us new colour spaces, wider gamut support (hello HDR), and the incrediblecolor-mix().Browsers are under constant development. Users might not see it, but developers certainly do.
There are people who build websites but aren’t full time web developers and aren’t even aware of any of that. We barely know CSS. Tailwind is popular because it makes knowing CSS someone else’s problem.
We certainly won’t care about color spaces. HDR? Yawn. The web was already colorful enough.
I appreciate the point, but I can't say I agree, Brian. Tailwind provides a lot of value, but it's not because it saves you from learning CSS. You still need to understand layout models (flow, float, flex, grid, et al). You still deal with inheritance via the cascade. You still need to understand specificity. TW isn't a deep abstraction over CSS.
I would argue that Tailwind is primarily popular because it's exceptionally well-suited to component-based design, provides standardization among teams, and is designed for build pipelines with features like tree shaking. It's the evolution of earlier approaches like BEM to writing tightly-scoped components without unexpected side effects everywhere. Atomic classes just integrate better into modern JS stacks like React and Vue.
I don't disagree that there are tons of people building pages using whatever tools they have available. Many of them are creating sites for their small business, or just making a homepage for their dog. Though I suspect most of them are using apps like Webflow and Wix, rather than CSS frameworks like Tailwind or even Bootstrap. So while they may not use these new CSS features directly, the tools like Wix they rely on are using them under the hood.
I think what we're really discussing here though is if these kinds of features should be moved from the page-level to the browser-level. A lot of features or complex layouts were possible in the past, but they might have required JS shims, workarounds, or other hacks. Browser vendors typically look at what people are trying to accomplish but can't, or what they do implement but in a clunky or poor way. Then they review solutions to make that better.
While I get the concern about making browsers too complex, I also think it makes sense to put the complexity in the browser, and not leave it up to every single page to implement in their own, non-standard way. Browsers are constantly evolving to optimize earlier functions, and even new functionality and enhancements can be added along the way. Some examples are
alert()windows adding an "ignore future alerts" checkbox, text areas becoming resizable, and form submissions adding retry prompts. Accessibility also plays a large role, as new features like<dialog>can be properly contextualized by screen readers rather than trying to guess at their meaning.In my opinion, websites are getting larger, but they should be getting smaller as popular features require less code to implement. Today I can create an effect where clicking a search button smoothly expands an input bar to fill the available space in under ten lines of CSS. Something like that would have required far more expensive JS and layout logic in the past. Websites being bloated and slow to load comes largely from wasted code, over-engineering, and poor priorities during development. It's not because browsers are too capable.
That said, I do think some bad APIs have made it into the browser. Web components are a bit faff, and IndexedDB was a bust. Arguably, the entire cascade model is not very good. But most APIs are well-researched and agreed upon via consensus, and are targeted to reduce specific developer pain points. All of the additions I described above were great candidates.
So when is it enough? That's a fair question, but realistically I don't think the web will ever be "done". It constantly evolves to meet new needs. Before 2007, nobody browsed the web by panning on touch screens. Yet overnight, the entire idea of a "cursor" disappeared, and we needed to rethink layout, scaling, and interaction. We addressed that with features like media queries, image
srcset, and multi-touch, but the effort is still ongoing.Who can predict what's next? What does the web look like in AR or VR, where we have gestures but no keyboard? When navigating through voice via wireless earbuds? By engaging with an AI agent? Or browsed via brain-computer interface? Interesting questions, and some we may need to answer eventually.
I think we can agree that the web has never been static. From the earliest support for tables and images, it has grown to meet our needs. As long as people continue to want to create and share experience online, I expect that will continue. Whether that means building something complex, or just making a homepage for our dog.
I do think the web standards folks are doing good work, and there are nice improvements. I'm not against that work continuing. I'm just saying, stagnation wouldn't be so bad.
If you had to build an ordinary website using only API's available five years ago, it seems like it wouldn't be too painful?
Five years ago wasn't so bad. Ten years definitely had pain points. Twenty years ago, oh boy.
So I can take your point that all the "major issues" have been worked out, and slowing the pace might be okay. But I also think it's the kind of assessment we can only really make in the context of the present. Twenty years ago, I thought it wasn't so bad using
spacer.gifimages to control table layouts. Twenty years from now, will I equally look back on today as the "dark ages"?For example, I wouldn't be surprised if we consider untyped JS to be reckless, cowboy programming. Or scoff at the amount of boilerplate required to add "simple things" like data bindings, in the same way we talk about jQuery now. But it's really hard to say without hindsight.
I'd also mention that it's not just about big changes. A lot of the recent focus has been on serving smaller groups, with features like enhanced accessibility controls, right-to-left formatting, better support for uncommon languages, and sensitivity toggles like
prefers-reduced-motion. With the web being such a fundamental aspect of our modern lives, it needs to be usable by everyone. A lot of work goes into making that happen.Realistically, I think stagnation only works for so long. If we collectively took a 5 year breather, then no, I don't think anything really bad would happen. But the problems would start to pile up. For users, features like passkeys would languish, because why use something that doesn't integrate well into your digital life? For developers, tools like style queries would only be partially implemented, and would be considered buggy/unreliable. And while you dismissed HDR colors before, apps like Pixlr need that to be able to work accurately with your new digital cameras.
I don't think stagnation kills the web, exactly, but it would certainly be less lively than before.
Yep, I agree. Also, there are likely lots of other improvements important to certain people or essential for certain specialized apps (like Pixlr) that would be unnoticed by most.
Agreed. See this problem and this proposed solution as an example. This would make building a small, personal site so much easier for amateurs and people who don't want to use JavaScript.
If web browsers are invested in less and fall behind, then companies will invest more in the closed ecosystems of iOS and Android apps instead. I don't think people appreciate the web ecosystem enough for being an open standards-based system with multiple interoperable implementations. Twenty years ago the ecosystem was bad and it was a crazy fantasy that it could ever get as good as it is now; I hope it doesn't peak and taper off here now.
This seems kinda like concern trolling? I agree what more competition in the web browser space. But this feels like "oh y'all were so gung ho to go after Google's anti competitive payments to protect their search engine. Now you're gonna get you comeuppance!"
Sounds like the big browsers will be competing with the smaller competitors who didn't have those kickbacks?