Live data from Hacker News

Apple confirms it's breaking iPhone web apps in the EU on purpose

techcrunch.com

701–710 of 829 posts

Re: Apple confirms it's breaking iPhone web apps in the EU on purpose

#701

Earlier quoted context omitted.

I love selfhosting/flashing as much as the next person, but why do you believe going AOSP would be the next step away from iOS? Someone moving of iOS should just pick up a Nokia device with a stock standard Android OS. It will serve them fine without all the hassle of flashing. If they want a higher end device with stock Android, go for a Pixel.

Because it's replacing one closed OS for another closed OS. AOSP is open source and unlike stock Android it doesn't have a bunch of preinstalled spying apps from Google/Samsung/you-name-it which you cannot easily uninstall. Plus Nokia is no more: https://www.gizchina.com/2024/02/01/mobile-phone-brand-trans...

A common user can not be expected to go through with flashing an Android device. A technical user, sure.

But a technical user could also simply go and disable much of the offending apps via `adb` effectively removing them (without uninstalling them) from the device. This is a much less drastic move than flashing a device.

And contrary to belief, Google is not monitoring everything that happens on the device. Let's give the aluminum hat a bit of a rest here.

Re: Apple confirms it's breaking iPhone web apps in the EU on purpose

#702
It's worth noting that Android solved this problem a long time ago. Home screen shortcuts just open in whatever browser you created the shortcut with. If apple hasn't fixed this a year from now, it'll be a clear indication that they are taking this as an opportunity to disadvantage the web platform.

Re: Apple confirms it's breaking iPhone web apps in the EU on purpose

#703
post #334

Earlier quoted context omitted.

> you can see that unless at an OS level you provide additional 'meta-security' features that allow apps that run in other apps to have fine grains access control that is managed by the OS, it's pretty much "security? Well, whatever...". I don't think that's the only solution. A simple alternative is to declare that "apps that run in other apps don't get to do anything at all." I.e. in this case, in response to a EU…

This would run foul against the DMA, unless they make safari PWA also less capable.

That’s what I meant/said — they’d neuter the PWA framework itself, which would mean that any PWA (including Safari PWAs) would just become “regular webpages but standalone.”

Re: Apple confirms it's breaking iPhone web apps in the EU on purpose

#704
post #334

Earlier quoted context omitted.

> you can see that unless at an OS level you provide additional 'meta-security' features that allow apps that run in other apps to have fine grains access control that is managed by the OS, it's pretty much "security? Well, whatever...". I don't think that's the only solution. A simple alternative is to declare that "apps that run in other apps don't get to do anything at all." I.e. in this case, in response to a EU…

Then everyone will just bitch PWAs can't do anything.

But only in the EU, and only on iOS. They'd still get enhanced capabilities elsewhere. (On iOS on any other continent; on Android anywhere; on ChromiumOS anywhere, or just Chrome on desktop anywhere; etc.)

And the nice thing about PWAs, is that there's no way for a PWA to know or care that it's being run "installed", and change its expectations/requirements — as there's just no web API for that. Instead, a PWA must just attempt to talk to each of these permission-gated APIs it wants to use, and find that it's now being [prompted for and] given access to them, rather than silently refused them.

So, unlike tightening the security model around regular native apps, tightening the security sandbox around PWAs shouldn't actually fundamentally break them — they should be designed to gracefully degrade when refused these capabilities. Presuming these PWAs were already ordinary fully-functional web-apps, which have just been progressively enhanced with these features when and where available, they'll just act like they do "on the web" — which should still deliver on the app's use-case. That's what the "Progressive" in "Progressive Web Apps" is supposed to mean!

Of course, some PWAs 1. will have been designed from the ground up as PWAs, and 2. will have a purpose/use-case that's very specific to the use of these high-integrity web APIs, such that they're completely useless without these PWA-only permissions. A video-chat PWA, for example, won't do much without access to your camera + microphone. There's no point to using these webapps as webapps — and often they don't even let you do so (i.e. they attempt to access the specific API they need on launch; if they succeed, they render the app UI; if they fail, they render a prompt to install the PWA.)

I don't know if you'd really call these PWAs, since there's nothing progressive about them — there almost needs to be a different term for these apps that need the high-trust APIs to do anything-at-all. For the sake of discussion, I'll refer to these as "Elevated Web Apps" (EWAs), since they require elevated permissions to be useful.

It's only these Elevated Web Apps that would benefit from having what the GP called "meta-security": the ability to interact with the OS security on a per-webapp basis, through e.g. an Android-like install-time gate where the app presents a capabilities manifest (displayed to the user as a set of permissions it wants) and the user makes a decision of whether to accept that.

And, if Apple simply neutered PWAs rather than removing them, it's only these Elevated Web Apps that people would "miss out on."

As cool as PWAs are as a technology, these Elevated Web Apps are a true minority or them — maybe 1% or so.

And — at least as far as I know — almost all Elevated Web apps only exist for one of two reasons:

1. to serve use-cases that users with access to native apps from an app store, just have no reason to care about. (Specifically, they were developed to allow users to accomplish native-app-equivalent things on OSes that don't support any kind of native apps — like FirefoxOS nee KaiOS, or early ChromiumOS.)

2. to benefit the developer at the user's expense, by forcing the user to give the developer permissions that allow the developer to spy on the user more effectively, before the app will work — but where the app doesn't actually do anything with these permissions to serve the use-case. (I've seen a few scammy Chinese dating sites demand to be installed as a PWA for this reason.)

In other words: on iOS, at least, you probably won't miss them! (Especially with the third-party App Store ruling also in place in the EU! Things like emulators don't need to be relegated to "WASM running in a PWA" any more; in the EU, they can just be third-party-store apps!)

Re: Apple confirms it's breaking iPhone web apps in the EU on purpose

#705
post #603

Earlier quoted context omitted.

That is simply just nonsense. They had 1,5 years from the time of being identified as gatekeepers to work on this. The DMA was voted on by the EU parliament and then the council in july 2022, Apple was identified as a gatekeeper in september 2022, the law became legally implemented in november 2022, with gatekeepers required to comply with it by march 6th 2024. I do not buy for a second that the richest tech company…

The simple answer is that it’s not worth it to them. They don’t see money with PWA at this point in time and therefore decided that breaking support was not a big deal. It obviously outrages everyone on HN, but HN is not your average customer of Apple.

Oh I don't care one iota about PWA's on iOS.

However the parent argument was a weak one, and so had to be answered with facts.

Re: Apple confirms it's breaking iPhone web apps in the EU on purpose

#706
post #693

Earlier quoted context omitted.

PWAs adhere to the same-origin policy, and all browser security policies associated. This means isolated storage (indexed db & local storage), isolated permissions, etc. Every modern browser has support for this. One webpage accessing the resources and data of another webpage is among the most basic of things globally known to be disallowed. This sandboxing reasoning is extremely bad faith.

Your argument hinges on a browser being a good citizen. The DMA makes no such requirement.

Apple could trivially audit browser apps and provide warnings if they do not adhere to basic security policies literally every browser has implemented. This is a basic, user friendly approach.

Their behavior is akin to a small, bratty toddler throwing a little tantrum, but instead of being a small toddler, it is one of the largest corporations on the planet. Their "little tantrum" impacts lives and livelihoods, because they are upset a population has reps that actually represent them.

I hope they get what they deserve.

Re: Apple confirms it's breaking iPhone web apps in the EU on purpose

#707

Earlier quoted context omitted.

> So Android allows alternative rendering engines besides Chrome for PWA? Yes: https://developer.mozilla.org/en-US/docs/Web/Progressive_web... I tested just now in Firefox with an app from https://appsco.pe and it does indeed work! I can do the same with the Android version of Brave. > If you install Firefox it uses Gecko but still has native app look feel? That depends on your definition. Making an app _feel_ native…

> I tested just now in Firefox with an app from https://appsco.pe and it does indeed work! I tested just now in firefox with an app from https://appsco.pe and it just...opened a browser tab with the website. So I understand a PWA is just a website but isn't the whole point to have a dedicated window/card for it?

I don't know what your setup is, but it did work for me, creating an app that shows as its own icon on the homescreen, without FF chrome, with a separate app-switcher entry. Using a S24 Ultra with whatever the current OneUI is.

Re: Apple confirms it's breaking iPhone web apps in the EU on purpose

#708

Earlier quoted context omitted.

I recommend that everyone interested in this topic read some of the comments from PWA developers at: https://bugs.webkit.org/show_bug.cgi?id=268643 Apple’s decision is going to kill businesses and break apps used by hundreds of thousands of people in Europe, many of whom are healthcare workers delivering patient care.

Patient care apps as PWAs? Yikes.

What's unusual with it? I even do my online banking exclusively via web browser.

Re: Apple confirms it's breaking iPhone web apps in the EU on purpose

#709

That's why I almost wouldn't buy a new Apple product anymore, with their draconian Chinese model (Apple as the big parent), and instead I crack Apple's product to assert my freedom (I do use Hackintosh, by the way). The only exception is an iPad Pro (M1) because there aren't good competitions in the market. Over the time I'm starting to think about replacing it with an Android tablet but I'm still yet to find one wit…

> Android tablet but I'm still yet to find one with a decent pen and memory

I don't have one, so take with grain of salt, but I've seen reviews of the Galaxy Tab s9 Ultra are pretty good [0]. Super expensive though. Also as a bonus, android is possibly getting a VM [1], so in the future you might be able to install full fat linux on high quality tablets.

[0] https://www.youtube.com/watch?v=sl0UUhmaiDU

[1] https://source.android.com/docs/core/virtualization/architec...

Re: Apple confirms it's breaking iPhone web apps in the EU on purpose

#710
post #91

Earlier quoted context omitted.

Apple has a decade+ track record of making devices that i really like. (At several points I’ve compared solutions across the market). Instead of siding with Apple, why would I side with anonymous and random internet commentators who have never made devices I want to buy?

It is definitely odd to outsource your moral principles based on which mega corp you opened your wallet to.

Morals? You think you’re some kind of righteous crusader?

I’m talking about toys and gadgets. The ability to view memes.

Post reply on HN