I have no wish to defend Apple. However, I have no wish to defend Google or Microsoft either. Collectively, those three have a monopolistic market share for operating systems and web browsers on both desktop and mobile.
My fundamental problem with this author is his massive conflict of interest. He's not an outside observer but rather a Chromium engineer, former Google employee, current Microsoft employee. He talks about "competition" and "competitors" while basically ignoring the monopolistic landscape of the industry and the role of his own employers in that monopolization. Nobody has clean hands here, not Apple, not Google, not Microsoft. I don't see any of them really acting in the best interest of consumers. Let's not pretend, for example, that Chromium doesn't push a bunch of shit that consumers never wanted.
Progress would be breaking up this triopoly, not allowing Blink/Chromium to dominate everything.
The web "standards" bodies are a joke now because of the dominance of these few companies over web browsers. I don't even want to hear about standards anymore. So-called standards now are just the monopolists coming to agreement among themselves. All we have here is the employee of one monopolist complaining about another monopolist.
As far as I'm concerned, the web standards should be so simple that a little indie developer could write a full-fledged web browser. But people like the article author want web browser engines to become entire operating systems, which in effect excludes almost everyone from writing a web browser. That's not openness and freedom. It's inherently monopolistic.
1. "Lil' indie dev can make an engine" is irrelevant in terms of standards or browser/engine marketshare unless he can obtain significant browser/engine marketshare. He can't. So framing the current state of the browser engine market/"monopolies" as a barriers-to-entry problem with engine/standards complexity is completely nonsensical. Especially when engines are fully open-source.
2. Super-simple engines that any indie can solo-code means disregarding 95% of feature demands by web developers (your audience, remember?), even basic things like encryption, video/audio streaming, chatting, conferencing, that users (your "actual" audience) expects and demands. It means killing basically every part of the web that isn't text documents, which means turning everything else into native apps beholden to their own monopolies, that are mostly crappy (contractor-made, less sandboxed/secure than on the current web). That's also nonsensical, if your objective is "consumers". And it means making every company's revenue dependent on Apple/Google/MS instead of a web domain they control.
Really you should examine your assumptions and logic far more thoroughly.
> "Lil' indie dev can make an engine" is irrelevant in terms of standards or browser/engine marketshare unless he can obtain significant browser/engine marketshare. He can't.
How is market share relevant?
What would be wrong with designing simple web standards and as a result having 1000 options each with 0.1% market share?
> Super-simple engines that any indie can solo-code
This is a straw man.
> video/audio streaming
There used to be browser plugins for things like this.
> web developers (your audience, remember?)
> users (your "actual" audience)
Which is it?
Catering almost exclusively to web developers, as the browsers have done, unfortunately, is not in the best interest of web users.
> native apps beholden to their own monopolies, that are mostly crappy (contractor-made, less sandboxed/secure than on the current web)
This is false.
> And it means making every company's revenue dependent on Apple/Google/MS
This is true regardless, native or web.
And people like the article author like to claim that Apple is forcing companies to write native apps, but notice that on Android, where there are alternative web browser engines, these same companies still write native apps. That's not because WebKit is bad, because WebKit is irrelevant on Android. Rather, it's because smartphone browsing is generally inconvenient and also native apps offer some inherent advantages over web pages.
To be clear, monopolies are bad, both for operating systems and for web browsers. But turning web browsers into operating systems won't break the current monopolies, because it's the same companies in either case.
The unfettered market has failed. We really need democratic governments to break up the monopolies. Whether that will happen or not remains to be seen.
> instead of a web domain they control.
Do you even "control" a web domain? Web browsers can throw scary warnings over http and force you to adopt https certificates. Then the web browsers can decided which certificate vendors to trust or distrust. And all the web browsers have so-called "safe browsing" that can arbitrarily decide, with no warning and no resource, that your domain is unsafe.
Market share is relevant to your argument because it determines the standards that determine engine complexity.
> turning web browsers into operating systems won't break the current monopolies, because it's the same companies in either case
False. Linux is a perfectly great mainstream desktop option thanks largely to web apps (+Electron). Basically everything's available. How incredible is that? Or swap Linux with any other alternative.
The fact is you can make an indie engine right now. It won't support most of the web, but then again you don't want any of that to exist because of the additional barriers to entry for engines. So what's the problem? Make it.
In your world, somehow every platform vendor stops making web browsers, and we might have a bunch of browsers with 0.1% marketshare. How? Ludicrous. And proprietary browser plugins, really? So you're not looking to reduce complexity after all, then?
> Market share is relevant to your argument because it determines the standards that determine engine complexity.
I talked about two different things, (1) what is and (2) what should be.
(1) "The web standards bodies are a joke now because of the dominance of these few companies over web browsers". (2) "the web standards should be so simple that a little indie developer could write a full-fledged web browser."
You appear to be criticizing (2) by assuming (1), but (1) is the opposite of (2). The latter is an ideal, not the sad reality.
> Linux is a perfectly great mainstream desktop option
Linux is not a mainstream desktop option, because consumers can't walk into a computer retailer and buy a desktop running Linux. That's why Linux has a practically nonexistent consumer market share.
Of course techies can run Linux, though I wouldn't call Electron apps "great" by any measure.
> In your world, somehow every platform vendor stops making web browsers, and we might have a bunch of browsers with 0.1% marketshare. How? Ludicrous.
It was purely a hypothetical scenario. The point of the hypothetical scenario was to explain why market share shouldn't be relevant to standards (as I also explained at the beginning of this comment).
> And proprietary browser plugins, really?
They were in fact not all proprietary and are not necessarily proprietary.
What's wrong with modularity? Some software vendors can specialize in HTML/CSS, some in video, etc.
> Again I don't think you've thought it through.
You can disagree with me, but these continuing, unnecessary, personal, condescending comments are in violation of the HN guidelines. Please stop.
A "little indie developer" hasn't been able to write a web browser in several decades. An implementation of 2005-era CSS and JS would be a complex task surely requiring a fairly sizable team to implement.
The fact is, all three major browser implementations are open-source and that has allowed any company to come in and release or embed their own browser with minimal effort. Chrome/Blink is dominant but that is not due to technical barriers that make it difficult for other companies to ship their own browsers. In fact it is now easier than ever to do so.
> A "little indie developer" hasn't been able to write a web browser in several decades.
IMO that's a major problem.
> The fact is, all three major browser implementations are open-source and that has allowed any company to come in and release or embed their own browser with minimal effort.
And they're all beholden to Google's decisions, for example to wreck the web browser extension API. Of course you can fork the code, but forks become increasingly difficult to maintain as they diverge significantly from the original code. Google can make things extremely difficult for forks.
No, the worst offender (by far) does not get a free pass just because the world is not perfect.
All of the main contributors have corporate interest and are not acting out of the goodness of their hearts, yes. But Apple's level of anti competitive, vertically integrated, gaslighting bullshit goes above and beyond.
> does not get a free pass just because the world is not perfect.
Nobody is giving Apple a "free pass". I started by saying literally, "I have no wish to defend Apple." I also said, "Nobody has clean hands here, not Apple".
> But Apple's level of anti competitive, vertically integrated, gaslighting bullshit goes above and beyond.
I would note that the US Department of Justice is currently pursuing two monopoly cases against Google and suggested that Google should divest Chrome.
Eh... the author is not operating in good faith. Just for starters, this little jab, re web usb:
> This is far from the Hacker News caricature of "letting any web page talk to all of your USB devices."
The author is a meretricious liar. OSes do not randomly download code and run it. More importantly, unlike the author's employers (google and microsoft), they are not in the habit of allowing randoms to pay 1/1000 of a cent per install to run their spyware if you don't use adblock.
Nobody comparing what code gets run on an OS to what code gets run in a browser (eg every time you load a web page) is operating in anything approaching good faith. The risk exposures are nothing alike.
The one piece from this author that I found most persuasive is the analysis of mobile web vs app usage share. The web is practically dead on mobile. Users spend practically no time there, it's night and day compared to desktop. It's no wonder that developing countries, where desktops had less penetration and mobile dominates, have much less of an "open web" to speak of, and far more things are done in apps/siloes. If we accept the premise that the mobile web shows us one potential future trajectory for the web, it's looking grim. And if so, Google's approach of trying to make the "Web Platform" match native platform capabilities (notifications, for example) is far healthier for the web than Apple's approach of gating features exclusive to native apps (it almost sounds self-evident when I put it that way). Beyond all the arguments and talking point, look at mobile web vs app usage and you can see the end-result of Apple's vision for the web: it's dead.
It doesn't seem right to blame this on just Apple. In developing countries Android is more common.
The reason I think things are more locked down in apps/silos on mobile is part of a general trend to restrict more and more what people can do with their stuff.
See Microsoft requiring an online account to use Windows. UEFI. DVD/Blu-ray players not allowing you to skip ads. Nintendo being able to brick your Switch remotely. Cars sending your driving data to insurance companies.
Your examples have nothing to do with the web, but the author does criticize Google severely for hindering web apps early on to favor Android apps (before mostly changing tack with the rise of ChromeOS). When tech went mainstream in developing countries, it was on mobile, and the native app platform was more capable than the web platform therefore the true "web" didn't take off (even non-app text webpages). The fact remains that Google is the only one trying to fix that, and giving the web a chance to not die and be supplanted by native apps.
The web is being locked down too. See Google's proposals like the Web Environment Integrity API. Already, Google blocks sign-in's from 'embedded browsers' (webkit or electron-based browsers). The Encrypted Media Extensions API along with proprietary Content Decryption Module already lock out all but large company's from making a compatible browsers.
> Nintendo being able to brick your Switch remotely.
This was wildly exaggerated; Switches cannot be bricked remotely. Nintendo can ban you from online services, but that has been around... for decades.
The warning in the Terms of Service is saying that if you use unauthorized tools, there's no guarantee of compatibility. Updates as far back as the Wii had this warning.
> Without limitation, you agree that you may not (a) publish, copy, modify, reverse engineer, lease, rent, decompile, disassemble, distribute, offer for sale, or create derivative works of any portion of the Nintendo Account Services; (b) bypass, modify, decrypt, defeat, tamper with, or otherwise circumvent any of the [functions or protections of the Nintendo Account Services]. You acknowledge that if you fail to comply with the foregoing restrictions Nintendo may render the Nintendo Account Services and/or the applicable Nintendo device permanently unusable in whole or in part.
It's hard to feel sympathetic when literally everything above is illegal anyway, under the CFAA and DMCA. To me, this is a "we could brick your console if you do illegal stuff," which is nothing new (you can't steal an iPhone without Apple bricking it). This is also, presumably, why Nintendo believes they can get away with it - there's no way you can possibly sue without exposing yourself.
Also, note that this specifically targets Nintendo Account Services. This isn't you casually modding or even cheating; this is you hacking your console, and then using that entry point (your console's unique authentication certificates) to try and get into Nintendo's infrastructure. Even the people hacking Switch 1 knew that if you probe the CDN in a way even slightly unusual, you get permanently banned instantly.
I'm not saying I like it; I'm saying that it's irrelevant to any decent customer who isn't planning to be taking inventory of Nintendo's online endpoints.
We're arguably in a better place than we used to be, in some respects. Consider the Capcom CPS2 - an arcade cabinet that self-destructs when the battery expires or is removed for any reason. Now that's unthinkable and actually anti-consumer.
My point is... the Nintendo Switch 2 doesn't really stand out as exceptional in any regard compared to historical practices that we culturally already accepted.
If can receive a firmware update over the internet; you bet it can be bricked remotely, or otherwise upgraded to be capable of remote bricking. That applies to any and all devices, including those already sold. For that matter, I've lost track of all the times Linux distributions have done remote bricking, albeit by accident.
I'm curious: what are the web standards/specifications that webkit does not implement, efficiently or at all, that prevent all but niche or insignificantly important functionality?
For that matter, what critical new functionality has been introduced in the web standards/specifications over the last 20 years that was not possible to implement prior? My quick and uninformed take is just: video.
As a casual web developer, it feels very much like the CSS, HTML, and Web API specifications are wellwell beyond what is critical and deep into the "specifications capture" phase of how companies compete.
While it's easy to dismiss some of these as niche, many of them such as file system access and USB device support would completely change how some software could be delivered.
By not supporting these, one is forced to make a native app in many cases. With support for some of these standards, entire classes of application could be delivered via cross platform web applications. Imagine being able to install firmware updates on USB devices just by visiting the manufacturers website, for one.
While we can argue about safety, the file system access problem can be solved in the exact same way apple solved it for native apps, with permission dialogs etc and only exposing the selected items when attempting to access resources on the device. This is indeed the approach Chrome/Edge use today.
If you look at the browser API specs Apple has chosen not to implement and Chrome/Edge has historically, it's hard not to feel Apple is trying to protect its own native app store at times. There are obvious self-interested reasons for Apple to protect it's own store/native app system; even "free" applications often generate revenue from in-app purchase transaction fees or subscriptions. If developers can more easily avoid shipping a native app, they can also more easily avoid paying Apple fees.
> For that matter, what critical new functionality has been introduced in the web standards/specifications over the last 20 years that was not possible to implement prior?
Wasm, 3D acceleration, PWA, geolocalisation, USB access, sandboxed local storage and offline mode, accelerometer and touch support.
If one thing, the web has declined in the past decade with a lot of things which could and should be on the web moving to phone apps where gatekeepers can take their toll because phone browsers are gimped.
Given that we’re in the information warfare era, it’s a given that Apple would be trying to sane-wash their anti-competitive practices (which they undoubtedly are), and drag their feet as much as possible on standards and interoperability.
This site has a category called "Browser choice must matter" which is nothing but a series of articles slating Apple and Safari. Nothing about Chrome's desktop dominance or Google's abusive practices at all :thinking-face:.
> Windows even asks you if you want to use something else other than Chrome if you're in the EU.
And not by choice. It was a 2010 legal decision. Pretty confident that if the court didn't force Microsoft to do it they wouldn't give you that option.
It will seem odd to most, but my big concerns about Chrome relate to Android and ChromeOS. In neither environment did Chrome win share competitively. I think this has made them weaker and less useful, and that was mirrored in the tremendous difficulty we had in getting expansions of web capabilities done within the Chrome team, nevermind what I am documenting in this most recent post.
Sadly, the CrOS problem will be partially resolved when Google trashes it with Android rebasing in upcoming releases. On the Android side, Google is still withholding WebAPK support from competitors (suppressing PWAs on Android) and has failed to follow Apple's lead on hotseat browser replacement when choice screens are shown in the EU.
But neither of the bad effects are nearly as structural or impactful as Apple's out-and-out suppression of the web on mobile, because wealthy people carry iPhones and they have all the power:
Because it's specifically about mobile browsers and not desktop as indicated in the category header and you can actually use a competing browser on Google operating system? :thinking-face:
I have no wish to defend Apple. However, I have no wish to defend Google or Microsoft either. Collectively, those three have a monopolistic market share for operating systems and web browsers on both desktop and mobile.
My fundamental problem with this author is his massive conflict of interest. He's not an outside observer but rather a Chromium engineer, former Google employee, current Microsoft employee. He talks about "competition" and "competitors" while basically ignoring the monopolistic landscape of the industry and the role of his own employers in that monopolization. Nobody has clean hands here, not Apple, not Google, not Microsoft. I don't see any of them really acting in the best interest of consumers. Let's not pretend, for example, that Chromium doesn't push a bunch of shit that consumers never wanted.
Progress would be breaking up this triopoly, not allowing Blink/Chromium to dominate everything.
The web "standards" bodies are a joke now because of the dominance of these few companies over web browsers. I don't even want to hear about standards anymore. So-called standards now are just the monopolists coming to agreement among themselves. All we have here is the employee of one monopolist complaining about another monopolist.
As far as I'm concerned, the web standards should be so simple that a little indie developer could write a full-fledged web browser. But people like the article author want web browser engines to become entire operating systems, which in effect excludes almost everyone from writing a web browser. That's not openness and freedom. It's inherently monopolistic.
1. "Lil' indie dev can make an engine" is irrelevant in terms of standards or browser/engine marketshare unless he can obtain significant browser/engine marketshare. He can't. So framing the current state of the browser engine market/"monopolies" as a barriers-to-entry problem with engine/standards complexity is completely nonsensical. Especially when engines are fully open-source.
2. Super-simple engines that any indie can solo-code means disregarding 95% of feature demands by web developers (your audience, remember?), even basic things like encryption, video/audio streaming, chatting, conferencing, that users (your "actual" audience) expects and demands. It means killing basically every part of the web that isn't text documents, which means turning everything else into native apps beholden to their own monopolies, that are mostly crappy (contractor-made, less sandboxed/secure than on the current web). That's also nonsensical, if your objective is "consumers". And it means making every company's revenue dependent on Apple/Google/MS instead of a web domain they control.
Really you should examine your assumptions and logic far more thoroughly.
> "Lil' indie dev can make an engine" is irrelevant in terms of standards or browser/engine marketshare unless he can obtain significant browser/engine marketshare. He can't.
How is market share relevant?
What would be wrong with designing simple web standards and as a result having 1000 options each with 0.1% market share?
> Super-simple engines that any indie can solo-code
This is a straw man.
> video/audio streaming
There used to be browser plugins for things like this.
> web developers (your audience, remember?)
> users (your "actual" audience)
Which is it?
Catering almost exclusively to web developers, as the browsers have done, unfortunately, is not in the best interest of web users.
> native apps beholden to their own monopolies, that are mostly crappy (contractor-made, less sandboxed/secure than on the current web)
This is false.
> And it means making every company's revenue dependent on Apple/Google/MS
This is true regardless, native or web.
And people like the article author like to claim that Apple is forcing companies to write native apps, but notice that on Android, where there are alternative web browser engines, these same companies still write native apps. That's not because WebKit is bad, because WebKit is irrelevant on Android. Rather, it's because smartphone browsing is generally inconvenient and also native apps offer some inherent advantages over web pages.
To be clear, monopolies are bad, both for operating systems and for web browsers. But turning web browsers into operating systems won't break the current monopolies, because it's the same companies in either case.
The unfettered market has failed. We really need democratic governments to break up the monopolies. Whether that will happen or not remains to be seen.
> instead of a web domain they control.
Do you even "control" a web domain? Web browsers can throw scary warnings over http and force you to adopt https certificates. Then the web browsers can decided which certificate vendors to trust or distrust. And all the web browsers have so-called "safe browsing" that can arbitrarily decide, with no warning and no resource, that your domain is unsafe.
Market share is relevant to your argument because it determines the standards that determine engine complexity.
> turning web browsers into operating systems won't break the current monopolies, because it's the same companies in either case
False. Linux is a perfectly great mainstream desktop option thanks largely to web apps (+Electron). Basically everything's available. How incredible is that? Or swap Linux with any other alternative.
The fact is you can make an indie engine right now. It won't support most of the web, but then again you don't want any of that to exist because of the additional barriers to entry for engines. So what's the problem? Make it.
In your world, somehow every platform vendor stops making web browsers, and we might have a bunch of browsers with 0.1% marketshare. How? Ludicrous. And proprietary browser plugins, really? So you're not looking to reduce complexity after all, then?
Again I don't think you've thought it through.
> Market share is relevant to your argument because it determines the standards that determine engine complexity.
I talked about two different things, (1) what is and (2) what should be.
(1) "The web standards bodies are a joke now because of the dominance of these few companies over web browsers". (2) "the web standards should be so simple that a little indie developer could write a full-fledged web browser."
You appear to be criticizing (2) by assuming (1), but (1) is the opposite of (2). The latter is an ideal, not the sad reality.
> Linux is a perfectly great mainstream desktop option
Linux is not a mainstream desktop option, because consumers can't walk into a computer retailer and buy a desktop running Linux. That's why Linux has a practically nonexistent consumer market share.
Of course techies can run Linux, though I wouldn't call Electron apps "great" by any measure.
> In your world, somehow every platform vendor stops making web browsers, and we might have a bunch of browsers with 0.1% marketshare. How? Ludicrous.
It was purely a hypothetical scenario. The point of the hypothetical scenario was to explain why market share shouldn't be relevant to standards (as I also explained at the beginning of this comment).
> And proprietary browser plugins, really?
They were in fact not all proprietary and are not necessarily proprietary.
What's wrong with modularity? Some software vendors can specialize in HTML/CSS, some in video, etc.
> Again I don't think you've thought it through.
You can disagree with me, but these continuing, unnecessary, personal, condescending comments are in violation of the HN guidelines. Please stop.
> we might have a bunch of browsers with 0.1% marketshare. How? Ludicrous. And proprietary browser plugins, really?
Such a market works for cars (kinda), vacuum cleaners, hifi, golf clubs, burgers, synthesisers, sofas, buildings, DIY tools, dresses, shoes.
We also had more browsers back in the 00s, and into the early '10s.
So, yes. Really.
A "little indie developer" hasn't been able to write a web browser in several decades. An implementation of 2005-era CSS and JS would be a complex task surely requiring a fairly sizable team to implement.
The fact is, all three major browser implementations are open-source and that has allowed any company to come in and release or embed their own browser with minimal effort. Chrome/Blink is dominant but that is not due to technical barriers that make it difficult for other companies to ship their own browsers. In fact it is now easier than ever to do so.
> A "little indie developer" hasn't been able to write a web browser in several decades.
IMO that's a major problem.
> The fact is, all three major browser implementations are open-source and that has allowed any company to come in and release or embed their own browser with minimal effort.
And they're all beholden to Google's decisions, for example to wreck the web browser extension API. Of course you can fork the code, but forks become increasingly difficult to maintain as they diverge significantly from the original code. Google can make things extremely difficult for forks.
No, the worst offender (by far) does not get a free pass just because the world is not perfect.
All of the main contributors have corporate interest and are not acting out of the goodness of their hearts, yes. But Apple's level of anti competitive, vertically integrated, gaslighting bullshit goes above and beyond.
> the worst offender (by far)
This is disputable.
> does not get a free pass just because the world is not perfect.
Nobody is giving Apple a "free pass". I started by saying literally, "I have no wish to defend Apple." I also said, "Nobody has clean hands here, not Apple".
> But Apple's level of anti competitive, vertically integrated, gaslighting bullshit goes above and beyond.
I would note that the US Department of Justice is currently pursuing two monopoly cases against Google and suggested that Google should divest Chrome.
Eh... the author is not operating in good faith. Just for starters, this little jab, re web usb:
> This is far from the Hacker News caricature of "letting any web page talk to all of your USB devices."
The author is a meretricious liar. OSes do not randomly download code and run it. More importantly, unlike the author's employers (google and microsoft), they are not in the habit of allowing randoms to pay 1/1000 of a cent per install to run their spyware if you don't use adblock.
Nobody comparing what code gets run on an OS to what code gets run in a browser (eg every time you load a web page) is operating in anything approaching good faith. The risk exposures are nothing alike.
The one piece from this author that I found most persuasive is the analysis of mobile web vs app usage share. The web is practically dead on mobile. Users spend practically no time there, it's night and day compared to desktop. It's no wonder that developing countries, where desktops had less penetration and mobile dominates, have much less of an "open web" to speak of, and far more things are done in apps/siloes. If we accept the premise that the mobile web shows us one potential future trajectory for the web, it's looking grim. And if so, Google's approach of trying to make the "Web Platform" match native platform capabilities (notifications, for example) is far healthier for the web than Apple's approach of gating features exclusive to native apps (it almost sounds self-evident when I put it that way). Beyond all the arguments and talking point, look at mobile web vs app usage and you can see the end-result of Apple's vision for the web: it's dead.
> The web is practically dead on mobile.
How much of that is Google not allowing you to run uBlock origin on mobile Chrome?
The web without access to block lists (not just ads, but also others like annoyances) is practically unusable, especially on mobile.
It doesn't seem right to blame this on just Apple. In developing countries Android is more common.
The reason I think things are more locked down in apps/silos on mobile is part of a general trend to restrict more and more what people can do with their stuff.
See Microsoft requiring an online account to use Windows. UEFI. DVD/Blu-ray players not allowing you to skip ads. Nintendo being able to brick your Switch remotely. Cars sending your driving data to insurance companies.
Your examples have nothing to do with the web, but the author does criticize Google severely for hindering web apps early on to favor Android apps (before mostly changing tack with the rise of ChromeOS). When tech went mainstream in developing countries, it was on mobile, and the native app platform was more capable than the web platform therefore the true "web" didn't take off (even non-app text webpages). The fact remains that Google is the only one trying to fix that, and giving the web a chance to not die and be supplanted by native apps.
The web is being locked down too. See Google's proposals like the Web Environment Integrity API. Already, Google blocks sign-in's from 'embedded browsers' (webkit or electron-based browsers). The Encrypted Media Extensions API along with proprietary Content Decryption Module already lock out all but large company's from making a compatible browsers.
> Web Environment Integrity API
Dead proposal, two years ago.
> Blocks sign-ins from 'embedded browsers'
I would too; browser automation tools are a spam farm's favorite tool.
> Encrypted Media Extensions API along with proprietary Content Decryption Module
Previously it was just called Silverlight and Flash. The demand isn't going away; but EME+CDM is definitely more open than it used to be.
> Nintendo being able to brick your Switch remotely.
This was wildly exaggerated; Switches cannot be bricked remotely. Nintendo can ban you from online services, but that has been around... for decades.
The warning in the Terms of Service is saying that if you use unauthorized tools, there's no guarantee of compatibility. Updates as far back as the Wii had this warning.
https://www.reddit.com/media?url=https%3A%2F%2Fexternal-prev...
What's it's "meant" to say and what it says according to the law are two distinctly different things.
https://consumerrights.wiki/index.php?title=Nintendo%27s_May...
> Without limitation, you agree that you may not (a) publish, copy, modify, reverse engineer, lease, rent, decompile, disassemble, distribute, offer for sale, or create derivative works of any portion of the Nintendo Account Services; (b) bypass, modify, decrypt, defeat, tamper with, or otherwise circumvent any of the [functions or protections of the Nintendo Account Services]. You acknowledge that if you fail to comply with the foregoing restrictions Nintendo may render the Nintendo Account Services and/or the applicable Nintendo device permanently unusable in whole or in part.
It's hard to feel sympathetic when literally everything above is illegal anyway, under the CFAA and DMCA. To me, this is a "we could brick your console if you do illegal stuff," which is nothing new (you can't steal an iPhone without Apple bricking it). This is also, presumably, why Nintendo believes they can get away with it - there's no way you can possibly sue without exposing yourself.
Also, note that this specifically targets Nintendo Account Services. This isn't you casually modding or even cheating; this is you hacking your console, and then using that entry point (your console's unique authentication certificates) to try and get into Nintendo's infrastructure. Even the people hacking Switch 1 knew that if you probe the CDN in a way even slightly unusual, you get permanently banned instantly.
I'm not saying I like it; I'm saying that it's irrelevant to any decent customer who isn't planning to be taking inventory of Nintendo's online endpoints.
Whatever their thinking the result is the same you don't have ownership over the thing you bought.
In that case, we've never even showed up to the battle, there's been copy protection and ROM encryption since arcade cabinets.
https://retrocomputing.stackexchange.com/questions/24481/whi...
We're arguably in a better place than we used to be, in some respects. Consider the Capcom CPS2 - an arcade cabinet that self-destructs when the battery expires or is removed for any reason. Now that's unthinkable and actually anti-consumer.
https://en.wikipedia.org/wiki/CP_System_II
https://www.youtube.com/watch?v=vCtXZM8iG-o&pp=ygULQ2FwY29tI...
My point is... the Nintendo Switch 2 doesn't really stand out as exceptional in any regard compared to historical practices that we culturally already accepted.
Arcade cabinets aren't/weren't consumer products.
Nintendo switch 2 is a consumer product.
AFAICT, Nintendo can't brick GameCube, N64, SNES, NES, Gameboy etc. remotely.
If can receive a firmware update over the internet; you bet it can be bricked remotely, or otherwise upgraded to be capable of remote bricking. That applies to any and all devices, including those already sold. For that matter, I've lost track of all the times Linux distributions have done remote bricking, albeit by accident.
I'm curious: what are the web standards/specifications that webkit does not implement, efficiently or at all, that prevent all but niche or insignificantly important functionality?
For that matter, what critical new functionality has been introduced in the web standards/specifications over the last 20 years that was not possible to implement prior? My quick and uninformed take is just: video.
As a casual web developer, it feels very much like the CSS, HTML, and Web API specifications are well well beyond what is critical and deep into the "specifications capture" phase of how companies compete.
While it's easy to dismiss some of these as niche, many of them such as file system access and USB device support would completely change how some software could be delivered.
By not supporting these, one is forced to make a native app in many cases. With support for some of these standards, entire classes of application could be delivered via cross platform web applications. Imagine being able to install firmware updates on USB devices just by visiting the manufacturers website, for one.
While we can argue about safety, the file system access problem can be solved in the exact same way apple solved it for native apps, with permission dialogs etc and only exposing the selected items when attempting to access resources on the device. This is indeed the approach Chrome/Edge use today.
If you look at the browser API specs Apple has chosen not to implement and Chrome/Edge has historically, it's hard not to feel Apple is trying to protect its own native app store at times. There are obvious self-interested reasons for Apple to protect it's own store/native app system; even "free" applications often generate revenue from in-app purchase transaction fees or subscriptions. If developers can more easily avoid shipping a native app, they can also more easily avoid paying Apple fees.
> For that matter, what critical new functionality has been introduced in the web standards/specifications over the last 20 years that was not possible to implement prior?
Wasm, 3D acceleration, PWA, geolocalisation, USB access, sandboxed local storage and offline mode, accelerometer and touch support.
If one thing, the web has declined in the past decade with a lot of things which could and should be on the web moving to phone apps where gatekeepers can take their toll because phone browsers are gimped.
Given that we’re in the information warfare era, it’s a given that Apple would be trying to sane-wash their anti-competitive practices (which they undoubtedly are), and drag their feet as much as possible on standards and interoperability.
- sent from my iPhone
I just wish that Apple would make Safari available on Windows again, or that there was an easy way to have an up-to-date WebKit browser for Windows.
This site has a category called "Browser choice must matter" which is nothing but a series of articles slating Apple and Safari. Nothing about Chrome's desktop dominance or Google's abusive practices at all :thinking-face:.
Almost as if you can choose a browser on the desktop :thinking-face:
Windows even asks you if you want to use something else other than Chrome if you're in the EU.
> Windows even asks you if you want to use something else other than Chrome if you're in the EU.
And not by choice. It was a 2010 legal decision. Pretty confident that if the court didn't force Microsoft to do it they wouldn't give you that option.
Unfortunately not, but the fact is choice exists everywhere and is even encouraged in some jurisdictions.
The author was for a long time a developer on Chrome and now works at Microsoft on Edge. I would not expect them to lead an anti-Chrome crusade.
The author here.
It will seem odd to most, but my big concerns about Chrome relate to Android and ChromeOS. In neither environment did Chrome win share competitively. I think this has made them weaker and less useful, and that was mirrored in the tremendous difficulty we had in getting expansions of web capabilities done within the Chrome team, nevermind what I am documenting in this most recent post.
Sadly, the CrOS problem will be partially resolved when Google trashes it with Android rebasing in upcoming releases. On the Android side, Google is still withholding WebAPK support from competitors (suppressing PWAs on Android) and has failed to follow Apple's lead on hotseat browser replacement when choice screens are shown in the EU.
But neither of the bad effects are nearly as structural or impactful as Apple's out-and-out suppression of the web on mobile, because wealthy people carry iPhones and they have all the power:
https://infrequently.org/2025/09/apples-crimes-against-the-i...
Because it's specifically about mobile browsers and not desktop as indicated in the category header and you can actually use a competing browser on Google operating system? :thinking-face: