Hacker News new | past | comments | ask | show | jobs | submit login
A critique of Safari's influence on the web (httptoolkit.tech)
199 points by pimterry on July 28, 2021 | hide | past | favorite | 210 comments

So in a nutshell, the author's argument is that Safari is "killing the web" because apps written specifically for Chrome, may not run on Safari.

I find it cringeworthy to even have to explain the problem here. "The web" is not an app written for one browser. By the same logic, Firefox is also "killing the web" because it can't run Chrome-only apps.

The web is all mainstream browsers, and the web standards. Do you want to know what's really killing the web? Google subverting the standards process and unilaterally adding all kinds of random "standards" (like connecting to USB and Bluetooth devices, for some reason) as part of the web, then guilting other browsers into not following along.

Instead of critiquing Google for acting as if they singlehandedly own the web, web developers are parroting their line and enabling them.

Fucking embarrassing.

What a disingenuous summary of the article. I thought the points about persistent bugs in Safari were particularly pointed. Also most of the features Safari hasn't chosen to implement are supported by both Chromium and Firefox and are standards.

If you think standard APIs for screen orientation, pixel density, or touch events are some Chrome conspiracy, please do explain why.

The author points to many APIs supported by Firefox, so it's not just Google. In addition, they point out that many of the APIs that were "contentious" were added to Safari years after all the other browsers.

The author has already addressed many of the things in your comment, including Google doing whatever they want (they clearly stated that it is unacceptable).

Mozilla is basically forced to do whatever Google wants at this point. Firefox doesn't have the marketshare to influence standards, and if it doesn't maintain feature compatibility with Chrome, websites will drop Firefox support, and Firefox will loose even more marketshare and have even less influence.

Mozilla quite clearly didn't want to implement EME, for example, but they couldn't afford for Firefox to not work with Netflix!

Apple has more freedom than Mozilla precisely because Safari users can't just switch to another browser. And for as much as I find that abhorrent and anti-competitive... if Apple is going to retain so much power anyway, I'm at least happy they're using it to prevent Google's total dominance.

Pretty sure you wouldn't be happy with Microsoft taking this approach.

It's abhorrent period.

Is it any more or less abhorrent than Google blatant embrace, extend, extinguish approach to the open web?

To be clear, I'm not happy with Apple's approach to iOS. At all.

But Apple doesn't particularly care what I think, so I'm glad they're at least making moves which help preserve the open web, regardless of their motivations.

That rebuttal doesn't make sense. You're claiming that Mozilla is dragged into implementing bad standards by saying EME is a bad standard that they were dragged into, but Apple implemented EME before Mozilla did. The example standards the author gave have nobody arguing against them.

I wasn't making a judgement on whether EME was good or bad†. I was saying Mozilla didn't want to implement it, and was forced to, whereas Apple has more ability to push back against browser features they dislike.

As it so happens, Apple has no issues with DRM; their products are full of it.

† I do think EME is bad, but that's beside the point.

> Mozilla is basically forced to do whatever Google wants at this point.

Doesn’t most of Mozilla’s revenue come from Google?


1) you can decide to use other browsers besides Firefox if some sites don't work on it, but on iOS you can't use anything else (every browser still uses safari under the hood)

2) you can still use "chrome apis" even if you don't use chrome. More and more browsers use chrome's engine (see: Opera, Edge, Vivaldi etc.) and while I am not happy with that, it is becoming more and more commonplace.

What web developers think:

- User: "This site is broken. Fuck Safari, I want Chrome on my iPhone!"

What users actually think:

- User: "This site is broken. Fuck it, I'll go to another site."

Safari isn't THAT BAD that sites can't work on it. It's mostly minor differences, and lack of some "nice-to-haves" that are not crucial to UX (or in fact, their lack improves UX).

No one cares about your site that much to go to another browser just to browse it ON A PHONE. This is why I said, outside the web dev echo chamber, no one cares.

I've been around for IE6, and comparing Safari with IE6 is honestly absolutely ignorant and laughable. But I did also make my sites work fine with IE6 back in the time. That's our job.

Safari will move at its own pace. The only thing I care about it, is that it's secure, power efficient, fast and user friendly. I speak also as an iPhone user, because users matter more than developers and their endless whining about features.

> No one cares about your site that much to go to another browser just to browse it ON A PHONE

Because they literally CANNOT switch from Safari’s rendering!?

The fact that you wrote this with so much confidence (featuring all caps and gratuitous swearing) signals to me that you aren’t genuinely listening to the argument(a) at all.

It signals to me that you are only interested in tribalistic hyperbole.

Besides: “It’s fine for me therefore it should be fine for you” is a terribly unconvincing argument.

No you missed its point : even in Android, nobody it’s going to change browser if your site is broken. Users will just assume your site is broken. And to be fair, it’s the same thing on desktop browsers.

So, like it or not, even if iOS allowed other rendering engines on browsers, you would probably be stuck with supporting Safari anyway.

>nobody it’s going to change browser if your site is broken. Users will just assume your site is broken. And to be fair, it’s the same thing on desktop browsers.

Sorry but, history says otherwise:

1) Netscape users came across "works best in IE" banners and IE usage skyrocketed.

2) IE users came across "works best in Firefox/Chrome" notices and IE usage plummeted in favour of those better browsers.

Chrome didn't get to where it is solely based on marketing - it had some very strong advantages in the early years.

>you would probably be stuck with supporting Safari anyway.

I don't disagree there. But increased competition would force Apple to up their game. A win for everyone.

>but on iOS you can't use anything else

This is a good thing, you need a monopoly(iOS) to fight against a monopoly(Chrome).

The debate requires so e nuance, where Chrome and Safari are at both ends of the extreme. The Chrome team seems tk have the resources to churn out one API (proposal and implementation) after the other, some of them are obviously made just for Google products. Safari on the other end does not even seem to care for very prominent APIs that have been standardized with Mozillas Support.

Somewhere in the middle lies Mozilla. They need to stay competitive to chrome, but also want good privacy and most importantly lack the resourcing to just implement all of Googles whims. You don't just implement a Bluetooth API into your browser that works on Windows, mac and Linux!

I disagree with the premise. I'll do a quick analysis on why.

Google Chrome is Google's master plan to remove responsibility from the Operating System and put it into their full control. Of course, they have operating power to move forward 300 APIs at the same time. However, they do it with the main propose of governing the ecosystem where those APIs are pushed (the web).

The only way to be up to date with all those changes is power (operating money).

Firefox is barely able to catch up with all the updates to the standards... there is obviously a burden on what is needed to be implemented for a browser in 2020 vs what was necessary 10 years ago.

However, Apple argument is not about delivering value to their users via a browser, since that's an ecosystem they don't control. They deliver value through their macOS/iOS SDKs, which are ecosystems they fully own. Apple moves Safari as minimal as possible (regarding new APIs added) to make sure people can still browse, but while still incentivizing their own routes.

This kind of API strategies are everywhere the software touches. From Apple moving Metal forward instead of Vulkan, Microsoft with Directx12. Google with Flutter (towards cross-platform API UI dominance) and so on.

I don't think is a question of what the companies want, but what do their users use. As long as Safari is the dominant browser in Apple's ecosystem, they will make sure you can max out the value received through other ways that align more with their vision and proprietary APIs.

So, if you consider "Web" what Google is pushing on now, it's clear Safari is not pushing forward in that ecosystem. But none of them do it for the sake of their users, but their ecosystems

Those that embrace Chrome have literally helped Google with their masterplan to turn the Web into ChromeOS.

Safari, regardles of Apple's selfish reasons, is the last man standing.

So your point is that the web sound be less capable, and Safari is the last browser trying to cap the web's knees.

That seems to me like the point of all these recent articles.

That seem like an ideal situation. Which 'web' are you referring to? The standards compliant, semantically structured one, or the Google 'web'?

I'm not buying the dichotomy. Apart from anything browser-specific, JavaScript's weirdnesses ought to demolish this ideal as a real possibility. It's just a question of how broken, how crufty.

How about how open? Browser specific JS is a misnomer because blink and V8 have ~70% market share. This is a bad thing for the open web. It was a bad thing when it was IE6, and allowing chromium based browser to grow beyond 70% while Google are firmly in the driving seat is more dangerous.

So instead of helping an evil master plan to have an open standard operating system that’s federated, you’d rather help an evil master plan to have a locked down walled garden OS controlled by a signal vendor, not opened sourced, with a 30% tax on all purchases, and you aren’t even allowed to install whatever app you want or to repair it when broke using a third party?

At least you can fork ChromeOS, and apps can be hosted or ‘installed’ by any entity with the freedom to choose centralized control or not (as many school districts too)

They (Apple) implement the features they believe to be the most useful to the users, saying that a browser is killing the web, because for any reason they don't want to play catch up with another company's browser, is a little bit over the top. You can either develop apps with disregard of Safari compatibility, or just use APIs that are compatible with Safari. Safari performs better than Chrome in MacOs and consumes less energy. Regular MacOs users don't care about Safari supporting some CSS property , most websites people use, load just fine. As a dev, I agree with you, Apple should have Safari in a different place, and the API compatibility should be way better, but, saying that is "killing the web", dunno about that, this article seems like a hate piece.

If anyone is "killing" the web, it is web developers using shitty frameworks and gimmicks on their i9 development machines, resulting in multi-second load times on 90+% of devices in use.

You know what gimmicks web developers use? They use babel and other transpilers that add code bloat in the final asset since they can't use standard JavaScript because a browser doesn't implement certain apis. Or having to use third party shitty libraries as workarounds because again a browser decided they are not going to include functionality that has existed in all other browsers since 5 years.

While that may be necessary for complicated webapps, most sites display text, images and a some video. The tools to display images and text have been available since the 90s. The tools for video (without flash) have been available since HTML5, so a decade.

New recently introduced gimmicks only solve problems that were already solved. There is no reason that a site such as reddit should have to run scripts for 11s on page load, when a pretty much feature-identical site such as HN manages the same with 100ms of scripts running.

Modern webdev is, for the most part, creating new problems that didn't need to be solved. It's a job placement program, nothing more.

> There is no reason that a site such as reddit should have to run scripts for 11s on page load, when a pretty much feature-identical site such as HN manages the same with 100ms of scripts running.

The really disheartening thing about this dichotomy is realizing that all of this extra effort, code, and time are in service of the banality of tracking and ads.

If it was just tracking and ads it would be a lot more performant honestly.

The transpilers aren't really the problem. The real problem is developers adding in all sorts of libraries for every little problem, and each library added is adding tens or hundreds of kilobytes to the final assets size.

I certainly don't mean it to come off as a hate piece - I would really like Safari to do well, and I'm definitely not a fan of Chrome's approach.

My concern is that the approach you describe - if Chrome goes further & further forwards and Safari refuses to keep up or engage whatsoever and more sites start to use those features - eventually leads to a world where everybody _must_ to use Chromium to use the web.

If we get there, the whole browser ecosystem collapses, and the open web is in very serious trouble.

I would rather a world where Safari actively leads web standards, and ensures that they are designed from the start with a concern for privacy and battery life.

Except that we're seeing a world where folks are sometimes deciding to leave Chrome for Safari (or others).

Even in my non-web-dev circles of average folks, they're saying "I hate Chrome" and going somewhere else for their browser.

Even if Chrome "goes further and further forwards" it won't matter if it becomes the de-facto hated browser over time. Safari is banking on the fact that security and stability are more important than "some new flashy CSS" and it is possible they might succeed by following this strategy.

Yet Apple gave up on Safari on Windows and isn't present on Android phones. Unless Macs and iOS devices take over the world (which may happen years from now), Safari will continue to be outnumbered and pushed aside.

Depends where on the globe one is.

There are markets where iOS is pratically 0% and there are markets where it has almost all the market for itself.

So it depends to whom certain business want to sell.

Right now, it's the top browser engine by a long long way, and its market share is either stable (on desktop) or slowly increasing (on mobile).

I don't think either of our circles are representative of the wider world unfortunately.

> If we get there, the whole browser ecosystem collapses, and the open web is in very serious trouble.

I'm not so sure about that. It doesn't seem to me that users are demanding ever more features in their browser. I'm personally just interested in relatively dumb websites, and the only features that could improve the experience is those that improve speed (so I did see some features on the list I wish Safari added, but not many).

And I'm not sure that approach is even sustainable in the long term. How big should the collection of web standards be? Maybe it's come to a point where it's good that a browser holds back a bit? I think that makes it easier for Firefox to keep up with the common standard. It's clear that they're a few years behind on a lot of things, but if Apple really tried Firefox could be the one browser that was always behind, and since nobody is really encourage to use it by any big corporation, it could easily die. I make a point of supporting Firefox on my Windows devices (and it's my back-up browser for Mac as well).

I think the Web being such a gigantic standard that only one or two big megacorporations has the chance to make a browser engine for is much worse for the open web, than Safari lagging a few years behind.

Google (WHATWG) putting "HTML living standard", "URL standard", etc. as heading on their their phone-book-sized monstrosities and calling it a day, then leveraging their power over Mozilla to bless it, then putting pressure over others as the media powerhouse they are, thereby driving browser vendors out of business, can hardly be described as "standard" or other process of finding consensus.

> If we get there, the whole browser ecosystem collapses, and the open web is in very serious trouble.

I don't get this take. Everyone says if Apple doesn't catch up to Chrome then Google will control the web... But like... If Apple is implementing everything Google releases in Chrome to "stay up to date" then don't they already control the web?

Isn't Apple staying back and saying "Nah, we won't be implementing that stuff" making sure that Google doesn't control the web?

> I would rather a world where Safari actively leads web standards, and ensures that they are designed from the start with a concern for privacy and battery life.

We can assume Safari is leading.. but is Chrome following? :)

I think the issue here isn't that Safari is lagging behind in features. IE did that too, and we know where it is today. It is the fact that Apple does not allow any other browser (to use it's own engine) on it's platform.

I don't mind telling users that a particular software doesn't work on Safari and to use Chrome/Firefox. and Many of them will switch too.. if they could.

Chrome aims to be the new Flash. Apple rejected that “engine” first.

I was nodding along with this article until this point:

> There is no plausible option where other vendors stop these features coming to the web entirely. In many cases, they're already there. There's only a world in which they stop them reaching beyond Chrome's 65% market share (~70%, including all Chromium-based browsers, or ~80% for desktop).

If 2 out of 3 browser engines decide not to implement a feature out of principle than arguably it isn‘t part of the Web and if Google was serious about being a good web citizen they‘d disable features (or make them only available to ChromeOS) where they can‘t convince one other browser engine to implement it.

Google is using a similar strategy to what Microsoft used in the late 90s. The difference is that MS implemented new tags and features in IE without any standard. Google is taking care of manipulating the standard, and then singlehandedly implementing these new features in the standard even if nobody else cares about it. Then it manipulates their services to require those features, so Chrome is the only option.

This is behavior typical of large corporations and monopolies, because they have the resources to throw at the creation of more and more features, at the expense of smaller competitors.

Yes, it is bad behaviour, but yes, they're absolutely doing it anyway.

I wouldn't necessarily describe this as malicious, merely impatient (perhaps too generous) - they've been doing controlled origin trials and similar for web bluetooth for example since 2015, but the other browsers haven't engaged.

I do think if Firefox/Safari had serious counter-proposals today that would work for everybody and limit the standards to mitigate the concerns, the Chrome team would happily compromise in return for wider compatibility.

> I do think if Firefox/Safari had serious counter-proposals today that would work for everybody and limit the standards to mitigate the concerns, the Chrome team would happily compromise in return for wider compatibility.

I believe you‘re right but that leaves no room to say “we fundamentally believe this feature should not be part of the web”. Something I consider quite reasonable with regards to Web USB and Web Bluetooth.

Lets hope that the anti-trust judiciary has enough technical understanding to realise how restrictive and damaging to competition - Apple’s policy of not allowing alternative web browser engines on IOS is. They should force Apple to allow alternative engines to Webkit, thus enabling much stronger competition to Safari and to native Apps on Apple mobile devices.

> They should force Apple to allow alternative engines to Webkit, thus enabling much stronger competition to Safari and to native Apps on Apple mobile devices.

The ironic part here is that back in 2007-2013 Microsoft was sued by the European Union [1], and was forced to implement a downloader GUI that offers a randomized overview of available Browsers [2]; and their website/download links.

Back then Microsoft's defense was that an enduser could just install another Operating System on the hardware if they didn't want to be forced to use Internet Explorer.

And now, 10 years later, here we are, where Apple absolutely controls the Hardware _and_ the Software. Yet the European Union does seemingly nothing against it.

[1] https://en.wikipedia.org/wiki/Microsoft_Corp._v._Commission

[2] (with screenshots) https://www.zdnet.com/article/microsoft-fined-731m-by-eu-in-...

But when that happened didn't Microsoft have something like 95%+ of the global desktop operating system market share? Whereas Apple has nowhere close to that in mobile market share and likely never will. They are not a monopoly in the same way Microsoft was a monopoly.

With Apple you have viable alternatives from Android handset manufacturers: alternatives that are suitable for the average person on the street.

With Microsoft this was not the case. When that case started Apple were basically dead in the water on the desktop... and then you had Linux, which was not suitable for the average user (yes, I'm sure we can all cherry pick a story about somebody's grandparent being perfectly happy with Linux back in the day but these anecdotes, whilst fun and interesting, do not add up to meaningful data).

The EU Commission and several other competition bodies are currently studying Apple's anticompetitive practices. The reason it has avoided their action until now is probably due to the technicality of the issue and its deep ramifications. Hopefully the numerous recent articles about the matter will help the regulators understand it and act on it.

Well, probably also that technically Microsoft had a monopoly in desktop operating systems (apparently 97% of desktop computers in 2000 according to an article I found), whereas Apple’s smartphone market share in Europe is just less than 30%.

It’s more subtle because it depends what you consider to be the market.

Windows was the dominant operating system of the PC platform. But the openness of the PC platform wasn’t a problem. On smartphone, you are bound by the hardware to keep the OS sold with it. So it could make sense to consider the « iPhone devices » market instead of the « smartphone » market (same for Android).

Perhaps they realize that opening up the browser on iOS would disproportionally benefit Google and nobody else.

Not just Google - Many thousands of developers would benefit from being able to create web applications that can run on desktop and mobile and are not handicapped on Apple devices.

I'd rather take the side of the end users. Don't forget developers are end users too. Developers benefit from an open ecosystem, not the locked down silo'd web that Chrome is pushing for.

What is the locked down silo'd web that Chrome is pushing for? Safari is pushing for a crippled web that requires developers to pay an Apple app tax to reach users.

You're talking about "developers reaching the users", but that's from the perspective of building websites and webapps. What about from the other side? What if you're a developer that wants to control your own computer & your interfaces to the web?

It used to be a lot easier 10 years ago, federated protocols were more common, websites were forced to be more accessible in a more fragmented browser marketshare. It wasn't until Chrome started doing forced auto-updates (and the other browsers followed) that really pushed to a monoculture and led to more locked down websites. Developers loved that it made their work easier, but there is a cost of freedom.

I don't understand. You can still control your own computer's interface to the web. You can install extensions to disable JavaScript functionality or even build your own browser to do so. The only frequently used web browsing platform that doesn't allow this is iOS, as far as I know.

>Back then [...] an enduser could just install another Operating System on the hardware if they didn't want to be forced to use Internet Explorer

That does not make much sense, coming from Microsoft. Users were never forced to use Internet Explorer on Windows* and could always install other browsers freely. No need to switch OSs.

*The only time Microsoft imposed the use of Internet Explorer, as far as I can remember, was back in Windows XP, when Windows Update came as a web application.

The issue that is so often missed is that it wasn't the fact that IE was bundled as the default or that it had the largest market share. Very simply, the issue was how that market share was attained. Microsoft threatend to withdraw OEMs Windows licenses if the OEM bundled a competing browser, thus levereaging their near desktop OS monopoly to dominate web browsers.

Exactly, the problem was the illegal tying.

> That does not make much sense, coming from Microsoft. Users were never forced to use Internet Explorer on Windows* and could always install other browsers freely. No need to switch OSs.

I think your parent's history is wrong, but I'm pretty sure you couldn't avoid using Internet Explorer if you were on Windows. You could also use other browsers, but so much crucial functionality passed through IE that you simply couldn't opt entirely out of it. (For example, it couldn't be uninstalled.) I think that, at least for a while, Explorer (the file browser) was basically IE pointed at the local filesystem.

The EU looks at the health of the market and there is plenty of competition and Apple only controls a small fraction of it. Windows on the other hand was installed on like >90% of computers.

Ironically Apple ensuring that every browser on iOS runs webkit is one of the only things preventing Chrome from having a complete hegemony of the web.

If you actually want stronger competition the best thing you could do is separate Chrome from Google.

What funding mechanism would chrome have seperate from Google?

A good concern given how firefox/Mozilla was co-opted to some extent

The same one Firefox procures?

So an agreement with Google to have Google as the default browser of Firefox?

Doh, misread the question

If this rule had applied to Windows, there would be no Mozilla.

Or forcing Apple to accept real competition in the browser engine market and fund their team appropriately. Google puts literally several thousands of manpower into Chrome while Apple puts roughly one tenth into Safari. This is not acceptable.

Yeah, you're right. Maybe we should decimate Google's chrome team while we're at it.

Let's then also hope that the anti-trust judiciary understands that a closed-loop Web where the browser, the so-called "standards", the content/ads, and gatekeeping are made by the same party, is antithetical to what made the Web successful and got public spending behind in the first place. And that solving an economic problem of the software industry (nobody wanting to install your app, and nobody wanting to buy it either when everything is "free" as long you accept erosion of your privacy) isn't the job of the web, especially when it clearly is detrimental to the original purpose of the web as a medium for easy self-publishing.

Good point about the standards being made by the same party. It's also hard to imagine that w3c & Mozilla haven't had their motives affected at some level.

> They should force Apple to allow alternative engines to Webkit

I would be A-OK with that if the alternative engines provided the same privacy and security guarantees to me that Mobile Safari does (with penalties for breaching those guarantees.)

This is about forcing Apple to allow other engines. You can still choose not to use them. It follows that there is no reason that your choice of engine should be the default and only option for everyone that uses an iOS device.

Users can't make an informed choice about what browser engine they use. They don't understand what a browser engine is, and what impact it has on their security, battery performance and so on.

So no, you can't choose. When we say "you can choose" we need to realize that this is only good when you can make an educated, informed choice. If you can't, you need to be protected. And that's Apple's role.

Random choice is not a choice. Would you have your infant "choose" what drugs to take for headache for example? What criteria will they use? Pill shape, color, size and flavor. Well that's basically also our proverbial mom and pop picking a browser engine.

Furthermore browser engines have this peculiar habit of getting into everything. Open any app at all and it's probably using WebKit without you realizing it. What if the app uses some random outdated fork of Netscape, why not? Do you realize what "choice" you made by opening that app?


Yes, users can make an informed choice about what product to use or to buy. This is true for computers, this is true for mobile devices, this is true for any kind of software, and obviously this is true for web browsers. People have been doing it on all other OSes since web browsers have existed.

There is no reason to prevent competition in such a market, and hopefully regulators will become aware of the issue and act on it soon.

So your answer to all specific problems I stated is "I reject your world and substitute my own".

Not one reasonable and intellectually honest person expects that the average phone user out there even knows what a browser engine IS, let alone compare two of them, or know where it's used (when it's not even disclosed).

Keep in mind smartphones are even more widely used than "computers" in general. So I guess even 8 year olds now are expected to do a security analysis on the browser engine their game uses before playing.


I honestly thought showing how consumers are able to make informed choices in other and as technical markets like computers, mobile devices and other softwares categories, plus the fact that it is the case for web browsers on any other OS, would be two strong enough arguments to dismiss yours...

Probably more than 95% of car owners do not understand how the engine of a car works. So, following your reasoning, it would justify having no competition in the car market?

Thankfully most of the world chose the free-market economy model, of which free competition is a pillar. There is no reason for the web browser market to be an exception and justify obstructed competition like Apple does on iOS.

Consumers clamored for Flash engine, against their interests. Chrome aims to be the new Flash. iPhone rejected one, to the benefit of open web, and must be free to reject the other.

Are you also arguing that Apple and Microsoft should restrict you to safari and edge on your desktop to protect clueless users?

Just because the average consumer doesn't understand browser engines, doesn't mean that those who do should have that option removed. Hide it in developer settings behind three warning prompts for all I care but make the option available to those who want it.

And drop the ad hominem rhethoric please.

What you suggest actually can't work, it means applications have no control themselves what engine they run on. Browser engines can't just be a config setting, you're basically asking for chaos.

Also you should review what "ad hominem" means. Saying "no one intellectually honest would say 2 + 2 is 5" is not ad hominem.

Calling someone delusional on a completely subjective issue, for voicing an opinion definitely qualifies as an ad hominem.

That's not ad hominem. You have it backwards.

Ad hominem doesn't mean "don't say bad words about me and my opinions". Ad hominem would be disregarding an opinion not by discussing the opinion, but by discarding the opinion based on WHO said it.

It's in fact very hard to commit ad hominem against an anonymous person online who has said nothing about themselves. If they said "I have a history of delusions" and I said "therefore your opinion has no merit", that's ad hominem.

The fact that despite that you constantly hear people complaining about "ad hominem" online is just that much funnier. Strawman is another one that most people love to say, while having no clue what it means.


I don't know or care who said the above opinion. I care about their opinion and their opinion was delusional.

Thinking the world at large knows, cares, and can make "informed choices" about the browser engines used throughout their phone apps is delusional.

Also, there's nothing "subjective" about whether all smartphone users are web developers, and therefore happen to care about browser engines.

You called the argument delusional, and therefore it wasn't an ad hominem, so you're right about that.

But, "Delusional" is often associated with name-calling, so it sounds more like you're calling the person delusional, rather than what they said. Something like "This sounds delusional to me" would avoid that confusion.

I did call the argument delusional, yes.

But even if I decided to call the commenter names, that wouldn't be ad hominem (link above on details).

It's rude, it's uncivilized maybe, we can have qualifications like that. But it's not "ad hominem", because "ad hominem" isn't about "you insulted someone".

> Users can't make an informed choice about what browser engine they use. They don't understand what a browser engine is, and what impact it has on their security, battery performance and so on.

> So no, you can't choose. When we say "you can choose" we need to realize that this is only good when you can make an educated, informed choice. If you can't, you need to be protected. And that's Apple's role.

I'm getting strong deja vu to arguments over authoritarian nanny states in this.

Even if I accept the contentious value judgment that people need to be guided and protected by an entity that has strongly conflicting interests to them, I still don't think this is a strong argument as the default browser would still be Safari. People with little to no understanding of what a browser is are hardly going to download some outdated fork of Netscape.

> Users can't make an informed choice about what browser engine they use. They don't understand what a browser engine is, and what impact it has on their security, battery performance and so on.

why do they have to make this choice, why can't they just use safari?

> Users can't make an informed choice about what browser engine they use. They don't understand what a browser engine is, and what impact it has on their security, battery performance and so on.

My 80+ year old grandma made the informed choice to use Firefox on macOS. She can't use Firefox on her iPhone. She wants to use Firefox everywhere. She disagrees with you.

> You can still choose not to use them.

Not necessarily - if an app I am required to use for work chooses to use Chrome/whatever as its engine, for example, I don't have a choice there. Same for my bank / doctor / school / etc.

What kinds of penalties do you have in mind? Do you think Safari should be held up to the same standard other browsers are if they surpass it in security or privacy on any front?

> Do you think Safari should be held up to the same standard other browsers are if they surpass it in security or privacy on any front?


Those guarantees are currently provided by Apple and you seem to trust them. What would stop you from chosing alternative browsers from other companies you trust to provide the same guarantees ?

Apple were heavily relying on Google and Yahoo from maps, integrated web search, to weather data for a long time. At no point they asked you if you wanted to give your location to Yahoo.

Would you give a pass to Apple for how they handled privacy in the past and not to Mozilla who comparatively had only a few hiccups ?

Let's leave the goalposts where they are. Besides, I'm extremely confident that Apple only ever used Google for mapping in the early days of iOS/iPhone OS. In which case, if you took the time to read the ToS, it's there.

They stayed on Google Maps for 4~5 years until the writing was on the wall that it wouldn’t help them in the future (android being the elephant in the room). They still happily relied on whoever service they could rely on. Do you remember the facebook and twitter accounts setup directly in the system preferences ? Or right now they’re still sending data to Weatherchannel, until they complete Dark Sky’s integration I assume.

Service providers where mostly in the deal to get user data, and Apple’s push on privacy only started well into Cook’s tenure.

But you have trust in them in their current business position (it only was “the early days”). By that token I don’t think it’s unrealistic to assume at some point you’ll find other companies that are either new or show they changed enough to get your trust.

And they tell you this is happening. The links I shared illustrate Google and Mozilla doing things without asking or informing the user. Apple do indeed have form in this area too, battery/processor throttling springs to mind.

With regards maps, although what this has to do with your original question is beyond me, Apple informed the user very clearly on where the data is from and what data they were collecting, they still do. It was originally a combination of TomTom and OpenStreetMap, by the way.

Also, Jobs was pushing privacy in 2010 - "Privacy means people know what they're signing up for, in plain English and repeatedly. I believe people are smart and some people want to share more data than other people do. Ask them. Ask them every time. Make them tell you to stop asking them if they get tired of your asking them. Let them know precisely what you're going to do with their data."

I agree that companies can change. Big companies rarely do. I'm concerned that it's only a matter of time before Mozilla make the decision to switch to blink, which would be the end for them and the open web.

Interestingly, of the complaints made in this article, 6 are Editors Drafts, not agreed upon standards, or Working Drafts, again experimental un-agreed 'standards'. I'll concede that a working draft is close enough to being standardised. The rest, bar one, are implemented in WebKit. So the complaint in the article is that the WebKit project is holding up the web by not including Google's experimental features in a bid to say Apple should be allowing other, less secure rendering engines on a platform where users store a ridiculous amount of PII. It's being framed as 'freedom' - I'm sceptical. They'll be caterwauling about 'choice', but by enforcing 'choice', there is very real, and frankly sinister, risk that it will be lost irrevocably along with any semblance of an open web. Careful what you wish for.

Apple allows alternate browsers on iOS.

Good luck having bunch of lawyers figure out what the impact of a browser "engine" is.

And honestly good luck having regular users care about this, either.

The reason Apple won't have a problem with it is because the whole argument is just in the echo chamber of web developers who want their new shiny.

Wasn’t it the same issue with telecoms ?

I don’t remember my parents losing sleep at night because third party ISP couldn’t get the same access to copper lines. Yet we got meaningful regulation of that market.

I think as long as there’s big enough third parties losing money there is hope for change.

So which parties are "losing money" here?

Microsoft I’d guess ?

If Google were to lose android I wonder if they’d also push for more regulation on Apple.

I wish I could just say “Firefox”, but I’m afraid they would be small potatoes in this.

Microsoft recently abandoned their own browser engine and switched to Chromium. What are they supposed to lose again?


Ah yes, the Facebook Browser Engine... (?)

Google monetizes chrome. Google is an advertising company.

Apple monetizes Safari. They sell hardware and subscription services.

Facebook monetizes in part through these browsers. The changes to Safari hurt Facebook revenue.

In terms of competition law, the skins-over-webkit they do allow are plenty.

Anti-trust doesn't so much care about competition to improve web APIs. Because it's not really a market, what with no money changing hands etc.

In MS/IE/Netscape times, the idea of licensing browsers was still fresh. And, more importantly, browsers differed In more than just synthetic benchmark scores: IE had ActiveX plugins that weren't available for any browser or operating system, and those were widely used, discriminating against other OS, an actual market.

A similar situation would be Apple intentionally crippling Google Docs, or amazon.com. The changes to ads and tracking are the closest we've gotten in the last decade, and Apple has a pretty good case that they are doing so with the user's interest in mind.

Interesting article, and well researched.

I have to disagree with the view point of the article however. I view technology as serving users. The author makes a great argument that Safari makes life more difficult for developers. I believe that Safari makes life easier for users. The latest privacy features and reporting are great.

I love my Linux laptops where I run Firefox and Chrome, but for day to day work, writing, and consuming books and digital content, I rely on the limited ecosystem of MacBook, iPad, iPhone, and Apple Watch.

If companies or web developers really want to use features not in Safari, please do show me a message that warns me to not use your web site. The digital world is a big place, and we all get to make our own decisions.

> I believe that Safari makes life easier for users

Except when it doesn't. I use Safari iOS (because there's basically no choice since only their engine is allowed) and I regularly come across webms that don't load or that decode atrociously. It would not be that bad if only they would allow other browsers to ship their own engine.

> webms that don't load or that decode atrociously

But why are you blaming Safari for that instead of the site choosing to serve up that particular format rather than something more well supported?

because [WebM is a web standard](https://www.w3.org/TR/mse-byte-stream-format-webm/) so it's natural for developers to expect browsers to support it. If Apple didn't like it they could have pushed back before it was standardized. If people don't like Google dictating so much of web standards, they should be asking Apple to step up and be a bigger player in deciding those standards. It benefits everybody for there to be some amount of cooperation among these big players

> This document was published by the HTML Media Extensions Working Group as a Working Group Note. [...] Publication as a Working Group Note does not imply endorsement by the W3C Membership. This is a draft document and may be updated, replaced or obsoleted by other documents at any time. It is inappropriate to cite this document as other than work in progress.

Specifically not a standard?

https://w3c.github.io/mse-byte-stream-format-webm/ is better in that it's now an "Editor's Draft" but still says "This is a draft document and may be updated, replaced or obsoleted by other documents at any time. It is inappropriate to cite this document as other than work in progress." which strongly implies to me that this is far from a "web standard", no?

ah you're correct, I misunderstood the meaning of Working Note. It seems like Safari is adding support for WebM this year, but even so since it's still an editor's draft I think it's reasonable to expect websites to provide alternatives to WebM if unsupported. I guess that's why it was not included in the list of complaints in the origjnal article

> it's natural for developers to expect browsers to support it.

Well, one important part of the web developers job is to /not/ expect their production to work on every browsers. It’s even an important singularity of this job.

I see too much webdevs folks around me who conveniently work with Google Chrome on their 32 inch screen on a $3k+ workstation.

The web is meant to work everywhere and fail gracefully when it can’t. And this rule applies also if you despise a browser.

Make life easier for users at the expense of developers could very well be Apple's mission statement, and I say that as somebody who did a long stint as a primarly Apple developer. Sometimes the needs of devs and users are at odds, and when they are Apple sides with the users. I'm glad there's at least one company in the market that takes this view.

Can you tell me how do I not use Safari on iOS ? Some help appreciated!

You are conflating a browser (Safari) with a web rendering engine (WebKit). You can not use other web rendering engines on iOS, but you are free to use any of the 100+ different browsers that exist on iOS. Some of them are privacy respecting, some are not, some of them are ugly, some are beautiful, some of them are free, some of them are paid.

You don’t deserve the downvotes here. Chrome used to be based on WebKit now it uses a different engine Blink. It didn’t suddenly become a different browser.

I'm sure everyone here has heard more than enough of this, but Chrome is the one actively killing the web, and more importantly turning it into a rube goldberg monstrosity before it's eventual inevitable abandonment by every sensible human being.

Safari/webkit, as far as I can see, is merely trying to help. Perhaps not proactively, but by showcasing slight conservatism regarding which kind of bullshit bells and whistles get implemented. They're no knight in shining armor, and I shudder as I compliment Apple on anything at all, but they're doing what they can while still keeping their users satisfied.

If you care about the web and want to see it survive, build to Safari's standards. It's the least YOU can do.

Also everyone complaining about mobile Safari is leaving out the parts where it excels — speed and battery life. Safari is the best browser looking at actual websites.

Safari is the reason I bought an iPhone because using Chrome on a flagship Android phone compared to Safari was night and day. Pages would load in milliseconds not seconds and scrolling and zooming didn’t stutter.

I really don’t want Apple to lose this fight and live in a world where you can’t use non-chrome browsers without dealing with broken websites.

About the broken websites.

Firefox handles most of the site perfectly `without special treatments`. It you develop a website against it, it will about 98% work on chrome, and the reverse also hold.

The safari... on the other hand is a terrible s*t. If you try to develop against other browser. Your site will break on it about 4 in 5 even only use old feature that ie11 supports.

Its web standard support is a total joke in my opinion.

It isn't really a excuse to ship a browser that is so broken. Given the small mozilla handles mostly fine with it.

How on the earth a such giant company can't?

Google took WebKit and diverged. Why should Apple have to run after it? Just because they’re big?

Devs mostly failed to realize both were WebKit, targeted chromium and not WebKit, allowing breakage and letting Google get away with it. Devs is more giant than Apple. What’s devs excuse?

I mean they failed css2 (exists since 2000) features that exists since forever and bugged even before the divergence. Those features works on literally everyone includes ie 11 except safari and even against the spec apple provided itself. They never got fixed. There are tons of obvious ancient bugs on the safari exists are since day one and seems no one it fixing it(even they confirm it is a bug on their bug tracking site). Its supports site is a total joke. This is also what I believe why the divergence happened.

If a bug exists in half of browsers and spec don't really defined it, then it's my fault.

If a bug exists in exactly one browser and behaved contrary to the spec, then it's browsers fault. And there are way too many to count on safari.

Safari's inability to ship working features is what's saving the web? Actually I had not heard that one before.

"Standards aren't add-ons to the web. They are the web." -Apple's HTML5 Showcase, 2010

Apple had it right back in 2010, when they showed the world what was possible with the web as a platform. They took a principled stance, and made the web as a whole better because of it.

If you ship a browser (especially if you don't let users install competing engines) then you need to support web standards. You don't get to pick and choose. Standards are not optional add-ons. Apple's own words!

Except when standards happen to be ChromeOS features in disguise.

So Editors Drafts are 'Standards' now?

I don’t understand why safari gets so much hate. Personally, I prefer it over all of the other browsers I have tried:chrome, ie, edge, Firefox, brave, opera, etc

You can't even test on Safari without investing in Apple specific hardware. It's not cross platform and Apple doesn't help devs either. Even MS provider VMs so you could actually test their damn closed browser.

Even the open source Epiphany Web browser from GNOME can roll out updates to the core WebKit faster than Safari. It's not a 1:1 comparison, but it's the best way I can test WebKit from Linux even if Safari lags on releasing.

Until you've actually developed for mobile Safari, and had to deal with absurd bugs and anti-user problems that have been around for a decade, it's difficult to understand the scale of the problem.

Even if you ignore any of the new shiny APIs and features being ignored by Apple, the general state of web support by Safari is atrocious and forces developers into many one-offs specially for that browser. This is actively anti user.

There are a lot of bugs that only show up on Safari, and there's no way to test them without a Mac or a service like BrowserStack.

Supporting Safari takes up a huge amount of my time as a developer. That's why you see these articles.

And nobody cares much outside chrome-web developers, that’s why you only see these articles from them and not users.

I thought the most important part of the article wasn't about the missing API. ( Many of them are draft status )

It was the buggy implementation of a feature that go unfix for years [0]. It say it is supported only to find out it really doesn't. Or those one-offs, I think [1] is a much better take on the problem.

And as a Web page browser. I dont have much complain. But in terms of Web Apps.... there is where the fuss is all about.


[1] https://daverupert.com/2021/07/safari-one-offs/

100% spot on.

Lately, it feels like more and more news publicly attacking Safari appear in Hacker News.

Safari either isn't attacking, protecting or killing the web. It has a different purpose.

Apple's main objective is to sell services and devices. iPhones, iPads and their accessories. Also, Mac laptops/desktops.

As most of us know, they use their Apple Store, with a 30% cut for every transaction to sell their services. If a company doesn't like that they end up in court (Epic vs Apple) or getting special conditions under the table.

Safari is another piece of the iOS/macOS ecosystem to browse the web. They don't have any hurry to implement fancy features like Midi, Bluetooth, Notifications, proper PWA support, CSS rules, JavaScript. After all, you should have access to those things if you submit your app to their Store and accept their cut. They implement the features they find interesting or important and that's all. Even the release circle is slower than other browser implementations. Obviously, allowing other engines can hurt their services division and that's why the alternatives use the provided WebKit on iOS as enforced by the ToS.

As a user Safari is a bliss, but as a web developer is sometimes a pain.

Maybe we should stop supporting Safari on web applications which require non existing APIs with an encouraging message to try Chrome or protest to Apple if you're on iOS.

Yes. Let webapp developers have their Chrome-only "web" (what's left of it anyway), and us others the web we used to love for ... web stuff.

As freelancer/WFHer, I'm already using Chromium for the unavoidable Teams, G-whatever-its-called-now, Zoom, or other videochat/collaboration app anyway and can run these as sandboxed Electron app right now just as well. Or even better since these type of apps tend to require frequent updates of both the browser and loaded JavaScript code due to living on the edge of browser APIs, release-early-and-often, ads, "telemetry", or whatever.

While on the way out, please advertise script-heavy sites, preferably with an "application/html" media type for content negotiation and/or proper metadata for search engines, to spare me bulk loading tons of JavaScript when accidentally visiting your page just to leave on the spot anyway. It's appalling browsers had to disable basic HTTP features such as caching because everything can and will be used for fingerprinting. AFAIU, both FF and Safari deliberately chose not to implement the worst offenders in WHATWG's JS API landscape for that reason. Worse, any "cure" for fingerprinting is bound to benefit Google even more as the only ones to receive sensor data of each and every contact individuals make with public information.

"It's dead, Tim"

> As freelancer/WFHer, I'm already using Chromium for the unavoidable Teams, G-whatever-its-called-now, Zoom, or other videochat/collaboration app anyway

Tangential to the topic: is there any reason you could not use FireFox? I switched a while ago and was pleasantly surprised to discover all web apps work perfectly, including the Zoom webapp and Google apps (these always feel bloated no matter what browser I use).

I'm using FF all the time (except on iOS); in fact, right now as I type. I guess the reason I'm using Chromium for videochat and customer/work-related collaboration is so that I can close Chromium when I'm done for good ;) and continue browsing on FF for the stuff I care about with many windows and tabs open. Also, I can just nuke credentials for customer-related access when I no longer need those.

Maybe I'm burnt by having FF experienced getting slower over time with script-heavy pages (memory or resource leaks?) in the past, and don't want to loose my browsing sessions? It's a moot point anyway since I don't consider collaboration/videochat apps part of the web, so in the end why not have "web" developers target Chrome if they absolutely want to depend on Google. It'll always be more work to target multiple browsers, and developers/bosses will always find arguments to reduce the effort and suck that is browser development/testing. For me, that isn't what the (extant) web is about; there's a reason that a browser, like an MP3 player, used to be more like a decoder/media player rather than all things to all people. I found that, with few exceptions, quality of content I care about correlates inversely with the amount of JavaScript crap on a page/app.

> As a user Safari is a bliss

Having switched exclusively to safari over last 1 year, I couldn’t agree more.

They have nifty features like sms OTP auto fill that make transactions super smooth.

Chrome on Android has this as well as far as I can tell.

With a few small tweaks, this comment could have been about Internet Explorer and Microsoft a couple decades ago.

Well, with Internet Explorer the problem was simpler and quite worse.

It had around 90% of market share at its peak. So Microsoft had a complete monopoly.

They stopped caring about Internet Explorer since their business was selling Windows XP licences and PCs with Windows preinstalled. As Safari, IE was an afterthought, a nice thing to have to craft a more complete desktop environment out-of-the-box (like Windows Media Player, Microsoft Works...).

Luckily Firefox, Opera (Presto[1]), Safari and Chrome shattered the monopoly in the early 2010s.

Safari is just the secondary browser in the ecosystem and pretty much the only relevant after Chrome and its derivatives. This days the duopoly has been achieved by fast pacing new web features and vendor lock-ins instead of grinding browser development to a halt (I lived the days of IE6/7 hell).

[1] Dang I miss the good old days of Opera Presto, it was great... Vivaldi is the closest thing and I use it everyday.

It had a monopoly, but you could use different browsers. I didn't have to throw away my PC and buy a Mac to install one of the better alternatives. And the market talked: Microsoft, your browser is obsolete, I'm going to use something else.

With Apple here, it's different. We have a similar set of question as in the days of yore (Is Google bringing much needed innovation to the web, or are they pushing their own proprietary features to control the web? Is Apple the last sane browser vendor preserving privacy and being user-first or are they purposefully holding back the web to maintain their App Store revenues?) but the ability of the market to answer is stunted by the fact the browser is tied to the hardware. If it happens I am not happy with Apple choice, I have to change my $1000 smartphone, my $400 smartwatch, ... just to use a different browser.

Safari is the new IE6 from a web developer perspective.

This is all well and good, but I'm paying a premium to use their products. If the bottom line of your argument boils down to "protest to Apple", then yeah: what do you think people in the comments here are doing? Apple knows how to boil frogs, so you can't act surprised when people start protesting about how warm the water is.

I'll happily burn karma to get the point across: Safari is not good enough for my needs, and it's starting to feel like MacOS is also not good enough for my needs. As Mojave continues to lose support in mainline apps, I find myself avoiding my Macbook for most things. Got some photos to process? Eh, I'll do it in Linux. Got code to work on? Who cares, it probably won't run anyways. It's that kind of latent frustration that drives people away from the platform, so people should at least be aware of it.

This is a good take , instead of "attacking" Safari with clickbaity titles, maybe we can take action and do actual stuff(when possible) like the ones you mention.

> Obviously, allowing other engines can hurt their services division and that's why the alternatives use the provided WebKit on iOS as enforced by the ToS

This is almost certainly true now, but the rules predate their big services push. The main thing that prevents other browser engines is:

- Apps cannot run arbitrary code. So an emulator that bundles a fixed set of games is fine (providing the copyright is kosher), but an emulator that allows you to supply your own games is not. This of course also rules out executing JavaScript on web pages.

- iOS apps cannot mark memory as executable, i.e. no dynamic compilation. This alone would tank the performance of any competing engine even without the above restriction.

Lame rules, imo. I understand their origin, but it’s one of many things holding the platform back.

so what if keep seeing anti-safari stuff on HN? If it's in the news.... and the audience here happens to align with observations etc. But also, for every 1 Safari post there's 10 other ones hating on Google on the regular, so whatever

Apple nuked Flash from orbit.

Their balance sheet of naughty and nice web wise will be hard to put into naughty for quite some time.

A lot of the sneering in these comments focuses on a strawman web developer, making bloated sites with looming frameworks that expect every browser to behave like Chrome and are too lazy to write native code for every platform under the sun.

I've had enough of these junk takes. Let's talk about a real world scenario.

I've been writing a simple web app in vanilla JS, that relies on some well established browser standards to accomplish the following goals that I think we can all get behind:

1. How do I make the UI look nice?

2. How do I make it usable and accessible, especially to people who rely on browser focus to use a website?

3. How do I make it behave consistently across Safari/Firefox/Chrome?

4. How do I make it run as fast as possible?

I've worked hard to reach these goals in tandem, and for the most part, things are working (you can see for yourself at https://rezmason.github.io/wireworld-player/ ). The UI is built from standard form elements, there's no buttons made of <div/>s; screenshots taken in the different browsers very closely match; it runs very fast, though that's a WIP; and most importantly, every browser properly handled element focus.

Or so I thought.

Last week I realized non-Safari browsers set the focus on whatever element is clicked. Combined with Safari's lack of support for focus-visible, this causes the same user interactions in and out of Safari to result in completely different browser focus states, while still having identical display states.

In my opinion this is a worst case scenario for accessibility and usability. I hope we can all agree that any GUI ambiguity that can hide a bug for two weeks can equally derail the experience of a site visitor. But this ambiguity is the default, and I don't think it can be resolved without a workaround written in JavaScript. As the article above points out, these Safari quirks are long-standing open issues.

But my main point is this: by developing and testing my app in Safari first, I wound up making a worse user experience for my visitors. My choice of browser was an obstacle to half of my goals. I'm not shooting for the moon, here— my project is pretty barebones. It's just DOM and 2D canvas drawing in a single thread, for Pete's sake. (For now, anyway.) But it would be better executed if I'd primarily tested it in Firefox and Chrome, or if Safari was better stewarded.

>Safari isn't protecting the web, it's killing it

Saving the web means killing the modern WPA-related feature-creep, and Google's mono-control.

Safari helps with both.

Although Web development belongs to my activities, it should have stayed as it was in HTML 4.

Native apps with APIs are the best approach.

Except they don’t span multiple platforms and are locked behind app stores with 30% tax. There’s also no “view source” or “inspect” readily available for native so a substantial pathway for transfer of education, analysis, etc of apps disappears or is made opaque.

There were no web apps when I started into computing.

Reverse engineering minified JavaScript isn't the only path to learning.

Sure, but it's a much better path than disassembling memory by rom dump which is how I got into computing 35 years ago. I also had the benefit of controlling 100% of my own computer, from $0000 to $FFFF, without App Store DRM systems and non-repairable sealed and glued cases making it difficult for me to get under the hood of the silicon I bought and paid for.

Monopoly app stores with no opt-out for a device you paid $1000 should be straight up illegal.

>Sure, but it's a much better path than disassembling memory by rom dump

Better is arguable. It is, however, easier.

Which method produces better developers though?

Modern web apps are good. I prefer to

- edit my files in Google Docs (seamless sync)

- play games w/o ever downloading them

- do design in Figma

- swipe on Tinder w/o ever having to install or uninstall it

- browse code on GitHub

I want as much stuff as possible to be on the web. I dislike having to install applications locally.

On a less self-centered level, the web helps small businesses succeed. Do startups have the time to develop 3 different applications? I don't think so. The power of the web allows people who don't need to be developing native apps to avoid duplicating their work.

The web seems extremely important to small business innovation.

As for "large" / "bloated" sites. Features like new CSS properties seem unlikely to add more bloat or degrade accessibility.

If we left the web as hypermedia / documents only we would have killed its potential (figma / repl.it / so many others wouldn't have existed).

Safari isn't killing the web, trying to turn the browser into the defacto computing platform is.

The web is already the defacto computing platform.

Safari isn't killing the mobile web, 'app nags' are

> I do want to recognize that the Safari/WebKit team are working hard, and I do desperately want them to succeed! Chromium's domination is bad for everybody

I'm mainly a Safari supporter precisely because Chromium's domination is bad for everybody. I'm glad the author of this recognized that.

Funny how Brave keeps popping up in these articles, while peddling full page ads for risky financial assets.

I got a little curious on the statuses of these standards and went on a bit of searching.

- CSS contain (CSS Containment Module Level 2) - Editor's Draft[1]. First published in 2019. Not supported by Safari/WebKit.

- CSS offset-path (Motion Path Module Level 1) - Editor's Draft[2]. First published in 2015. Not supported by Safari/WebKit.

- CSS overflow-anchor (CSS Scroll Anchoring Module Level 1) - Editor's Draft[3]. First published in 2020. Not supported by Safari/WebKit.

- Resolution media queries (dppx) - W3C Recommendation since 2012[4]. Not supported by Safari/WebKit.

- :focus-visible (Selectors Level 4) - Editor's Draft[5]. First published in 2011. Not supported by Safari/WebKit.

- Touch Events - W3C Recommendation since 2013[6]. Supported by Safari/WebKit since 2010 (iOS 3.2). I assume the author meant Pointer Events[7] which became W3C recommendation since 2019, and supported since 2019 (iOS 13.2).

- BroadcastChannel - WHATWG Living Standard[8]. Blocked by privacy concern on WebKit side since 2020[9]. Initial support landed on WebKit trunk as of 2021-07[10].

- beforeprint/afterprint - WHATWG Living Standard[11]. Supported by Safari/WebKit since 2019 (iOS 13).

- Regex Lookbehind - ECMAScript 2018[12]. Not supported by Safari/WebKit.

- scrollIntoView (CSSOM View Module) - Editor's Draft[13]. First introduced as an update to CSSOM View Module in 2011. Not supported by Safari/WebKit.

- Screen Orientation API - W3C Working Draft[14]. First committed in wc3/screen-orientation in 2012. Not supported by Safari/WebKit.

- Date and time input types - WHATWG Living Standard[15], partial support by Safari/WebKit since 2012 (iOS 5) but no week/min/max.

- Service Workers - W3C Candidate Recommendation since 2019[16]. Supported by Safari/WebKit since 2018 (iOS 14.5).

- AbortSignal - WHATWG Living Standard[17]. Supported by Safari/WebKit since 2018 (iOS 11.3)

- Intersection Observer - W3C Working Draft[18]. First published in 2017. Supported by Safari/WebKit since 2019 (iOS 12.2).

- Client-side form validation - WHATWG Living Standard[19]. Supported by Safari/WebKit since 2017 (iOS 10.3).

[1]: https://drafts.csswg.org/css-contain/#contain-property

[2]: https://drafts.fxtf.org/motion/#offset-path-property

[3]: https://drafts.csswg.org/css-scroll-anchoring/#exclusion-api

[4]: https://www.w3.org/TR/css3-mediaqueries/#resolution

[5]: https://drafts.csswg.org/selectors-4/#the-focus-visible-pseu...

[6]: https://www.w3.org/TR/touch-events/

[7]: https://www.w3.org/TR/pointerevents/

[8]: https://html.spec.whatwg.org/multipage/web-messaging.html#br...

[9]: https://github.com/whatwg/html/issues/5803

[10]: https://bugs.webkit.org/show_bug.cgi?id=227924

[11]: https://html.spec.whatwg.org/multipage/timers-and-user-promp...

[12]: https://262.ecma-international.org/9.0/

[13]: https://drafts.csswg.org/cssom-view/#dom-element-scrollintov...

[14]: https://www.w3.org/TR/screen-orientation/

[15]: https://html.spec.whatwg.org/multipage/input.html#date-state...

[16]: https://www.w3.org/TR/service-workers/

[17]: https://dom.spec.whatwg.org/#abortsignal

[18]: https://www.w3.org/TR/intersection-observer/

[19]: https://html.spec.whatwg.org/multipage/forms.html#client-sid...

Try to load this in Safari, then try to load it on any other browser: https://netops.is/safari-push-bug/

That bug has been present for at least two major iOS version. I’ve reported it to Apple, but they’ve never bothered to fix it. On the plus side, it does load eventually now without crashing the browser, but the underlying bug is clearly still present, and it takes a long time to load.

Note: To test it in something other than Safari, you will need to use something other than iOS, because all browsers on iOS are wrappers around Safari.

Excellent article--thanks for it. California has the initiative process. I don't know how specifically to frame the issues, this issue. However I regard the web browser as pivotal in history as the invention of paper and the printing press. I also know large ride-hailing companies directly used the initiative process to undermine fair competition in California of all places. If you frame the issue properly it could have an impact. I recall seeing initiatives around caged chickens, which if passed hurt protein sources for single moms and poor families. And force feeding geese grain (which most animals regard like humans do heroin). So if you frame the issue properly you'd like get some results.

The reason I use safari is because it is the closest to a 'web browser', it doesn't have service workers, it doesn't support PWAs. The amount I have to turn off before it is usable is just less.

Safari has ServiceWorkers. https://caniuse.com/mdn-api_serviceworker

Frankly, I wouldn't care half as much about how bad Safari was if it wasn't my only option on many Apple devices. I had to use an iPad for school for a few years, and Safari is yet unmatched for "frustratingly pointless incompatibilities" in the browser sector, even compared against the dumpster-fire options like Edge and Explorer.

As a user, I love Safari for privacy and its reader mode which means I don't need to see ads. I switched to an iPhone due to privacy concerns with Android and haven't bothered installing Chrome on the new phone.

<img loading=lazy> is still not enabled by default iirc.

I'm super happy to see so many responses from people that realize Chrome is not a good thing anymore.

By the way, from 2008 Safari has MathMl support, while Chrome still hasn’t (almost 10 years passed after Google removed it).

I often wonder how many bytes of data and seconds of rendering time would be saved in the last decade if the most popular browser supported MathMl?

It sounds like safari is "destroying the web" because some devs would like features that would annoy users, or turn their websites into apps.

AV1 support would be nice though. Everyone should push for that.

Did it every occur to anybody that a breakneck race to turn the browser into an app runtime so complicated only the established players (ie Apple and Google, not even MS could keep up) are the only game in down might be killing the web (globally connected hypermedia library)?

Any of the big players could keep up if they wanted to, it's just that none of them have an incentive too.

* MS cares a little bit about the web because they have a few web apps that are pushing the envelope, but they really don't benefit from a powerful browser as much as Google does (at least yet). The #1 reason they have a browser is to push Bing.

* Amazon's websites (twitch, Amazon Prime Video..) could probably run on IE6.

* Apple's revenue stream would get hurt by people moving to the web.

* Facebook... well it's complicated. For now they've been doing a good job with their VR browser, but in the long term they will likely pull an Apple and make it not possible for VR website to be as good as the VR apps purchased through the oculus store.

(I don't know enough about all the big Chinese players)

Yeah yeah you keep the constant notification prompts, thanks.

The article is specifically talking about how this argument doesn't hold water, and ways that Safari can better engage with progress on the web so that we don't all have to use Chrome (complete with notification prompts aplenty) for everything in 5 years from now. It's not arguing that Safari should add notification prompts.

Many web developers I know constantly complain about Safari. While nowhere close yet, it's slowly inching towards the annoyance levels of IE.

It seems like the WebKit team was relying a lot on Google contributions, and the Blink fork (in 2013) was the starting point for a slow but steady decline.

Apple is swimming in money, yet they apparently can't staff their Safari team enough to keep up with Firefox, which notoriously lags behind Chrome. Not just feature wise but with stability and polish. Eg the flexbox implementation was extremely buggy for years.

I do believe part of it might be strategic to prevent webapps from becoming too useful on mobile, but that can't be the whole story.

Definitely putting on my fireproof suit here, but consider it possible that

(a) The web is a failure


(b) Chrome monopoly is the best outcome for the web.

OK, obviously the web is great, I'm using it right now, and it hosts millions of useful websites, so why is it a failure?

Consider the simple act of buying something. To make a purchase on the web, I have to type in a 16 digit credit card number, 3 digit CRC number, expiry date, name and address. Every single time. After 30 years of the web, nobody has figured out a way to make this securer or simpler. Well, no, I'm sure many have figured out a way, but nobody has implemented it.

By contrast, on my phone, I can buy with a simple fingerprint check. Apple provides that service, I find it valuable, and the firms in Apple's walled garden of apps find it valuable too.

Why is payment easy on iPhone? Because Apple, as monopoly provider of that walled garden, makes a profit from making it easy.

Why is payment hard on the web? Because there is nobody to make a profit from making it easy.

In short, the web's infrastructure is a massive public good, and like many public goods, in the absence of government coercion it is underprovided.

Web payment is just one example, and I'm sure there are others, but it's an important one. The lack of a decent payment infrastructure leads to pathologies. When paying for content is a hassle, people give away free content and use alternative monetization strategies like advertising - so we get Facebook, "you are the product" and manipulation by algorithms. When paying for items is a hassle, it's easier to just use Amazon who have stored your credit card already - so we get a huge ecommerce monopoly.

If Google effectively owned the web, change would happen faster; users would be more satisfied; web development would get easier instead of being a mass of horrible hacks around 30 years of legacy standards.

There'd still be competition. Apple, Google and Facebook would compete as platforms. That's an oligopoly, but platforms are natural oligopolies - nobody wants 1000 different platforms.

Conclusion: Chrome's dominance is probably efficient.

> To make a purchase on the web, I have to type in a 16 digit credit card number, 3 digit CRC number, expiry date, name and address. Every single time.

When I make a purchase in my browser, either:

* I select a card from the autofill drop-down and validate with CVC or fingerprint

* I'm already logged in, and the site remembers previous cards

Is it possible you have a bad experience here because you've asked your browser not to remember credit cards?

FF user. I don't see any option related to credit card usage. I didn't find anything in about:config either, though the defaults mention credit cards in some places.

The subtler point is whether a website owner can put in place a simple payment system - and be sure it will work on all clients. This is what you can do with Apple Pay, right?

Umm they can? What is PayPal and the like if not exactly that? I don't enter my CC every single time I buy online with PayPal. I just login and choose the card/address I want. The only thing Apple Pay offers is the ability to login with TouchID/FaceID.

Paypal is one solution, yes. Unfortunately, though I have paypal, most of the places I buy at don't offer it. There's a two-sided market issue. Sellers want to be sure that many buyers use Paypal before they commit to integrating it. Buyers want to be sure that Paypal is ubiquitous before they commit to having an account. In fact, neither of these conditions holds for everyone. So Paypal makes things easier, yes, but only for a minority of my purchases.

Is that not true of Apple Pay too? Lots of places I can't use Apple Pay. Amazon, Steam, Epic Games, etc.

I'd argue PayPal is way more ubiquitous than Apple Pay.

Some sellers don't want to use a service like Apple Pay/PayPal because they don't want to give up the money from the associated fees like Amazon for example.

There are lots of places on the web where you can't use Apple Pay, but on iPhone apps it is fairly ubiquitous. Amazon is an exception, perhaps unsurprisingly.

> FF user. I don't see any option related to credit card usage.


"currently available only for US and Canada Firefox releases." I live elsewhere.

> Consider the simple act of buying something. To make a purchase on the web, I have to type in a 16 digit credit card number, 3 digit CRC number, expiry date, name and address. Every single time. After 30 years of the web, nobody has figured out a way to make this securer or simpler. Well, no, I'm sure many have figured out a way, but nobody has implemented it.

> By contrast, on my phone, I can buy with a simple fingerprint check. Apple provides that service, I find it valuable, and the firms in Apple's walled garden of apps find it valuable too.

> Why is payment easy on iPhone? Because Apple, as monopoly provider of that walled garden, makes a profit from making it easy.

There are tons of competing companies out there that make a profit by making online payments easier for vendors and users. Vendors that support stripe allow you to make purchases within ~3 clicks.

Yup. If everyone used Stripe, I'm sure things would be better.

My point wasn't that there's no tech which makes payment easier. There is. My point is that there's no standard; as a result there are many different technologies and providers, none of which have dominance; as a result, I still have to fill in my cc details for the majority of websites where I purchase something. (Not for the majority of purchases - because I make most purchases on one website, Amazon. But then we're back to my point about pathologies.)

> To make a purchase on the web, I have to type in a 16 digit credit card number, 3 digit CRC number, expiry date, name and address. Every single time. After 30 years of the web, nobody has figured out a way to make this securer or simpler. Well, no, I'm sure many have figured out a way, but nobody has implemented it.

Is that really true? There are payment processors that fulfill the role Apple does when you pay on their device and, just like, Apple, make profit from those transactions.

Paypal may not be as dominant in the grand scheme of things (anymore?) making the market of transaction processing somewhat fragmented but you still don't need to enter credit card details every single time you buy from a new vendor. I certainly don't.

That's true. I only need to do it, say, 70% of the time. But on my phone, I only need to do it 0% of the time.

> Consider the simple act of buying something.

For a significant portion of the population (disclaimer: me included) this equals to going on amazon and pushing a few buttons. That’s on par with anything Apple provides.

I think your whole point about the web being complex is because you’re ok with that complexity so you engage with it. People who want to keep it simple naturally do so (we don’t gush over GAFA for no reason. A ton of users live their online life within these 5 companies).

"The lack of a decent payment infrastructure leads to pathologies.... When paying for items is a hassle, it's easier to just use Amazon who have stored your credit card already - so we get a huge ecommerce monopoly."

Sorry I didn’t address it directly; It’s a valid view in isolation, but you can’t complain about Amazon’s ecommerce monopoly and put forward Apple’s own restrictive solution as a counter to that. Both aim at keeping you captive in their ecosystem, neither are healthy by your standard.

That's fair, but these ecosystems are not the same. Many different shops have apps for iPhone, and Apple doesn't charge on purchases made from those shops. Amazon does, and it may choose to charge monopoly prices on certain items.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact