The EU is a total clusterfuck when it comes to tech regulation. Maybe the only good thing out of the EU is GDPR, everything else has been a massive overreach doing more harm than good.
Sound and fury is nothing. Actually fix some of these.
Google uses different methods (e.g. until fairly recently third party app stores couldn't auto-update the apps they installed, now many apps require attestation that fails if you don't run Google's code), but somehow or other the vast majority of Android phones still end up with Google Play and Google services.
Microsoft and Google both have cross platform frameworks for iOS and Android.
A translation layer between a generic API and Apple's isn't the same thing. It can't add things Apple's doesn't already have. For example, how do you run CUDA or OpenCL code on Apple's devices? Can a cross-platform framework provide a payments implementation so your cross-platform app can use a cross-platform payments system for in-app purchases? No, because they make you use theirs and apps that compete with theirs aren't allowed.
Visual Studio Code is clearly an app. Apple doesn't allow it on iOS. Apple makes a competing app -- for which they want you to buy a Mac.
You can't just use a definition of "app" that means "whatever is in their app store" and then claim that whatever they don't allow in their app store isn't an app.
Apple never stopped VSCode from being on iOS. There are plenty of IDEs for the iPad built by third parties.
No one would ever call a virtual machine an “application”.
Are you making up things now? You don’t run CUDA on any non Nvidia hardware.
But yes, you are free to port OpenCL to iOS if you want to.
> Can a cross-platform framework provide a payments implementation so your cross-platform app can use a cross-platform payments system for in-app purchases
A “payment API” is not an “application”. Yes the cross platform frameworks do expose the underlying API.
And you can have a cross platform app that forces you to subscribe outside of the App Store - Spotify, Netflix, Microsoft, etc all do that.
> No because they make you use theirs and apps that compete with theirs aren't allowed.
Are you saying that their no competitors on iOS to AppleTV, Apple Music, Apple Podcasts, Apple Books, the iWork suite?
CUDA drivers exist for pretty much every platform. They don't work without an Nvidia GPU, but in-theory nothing stops you from running CUDA code on iPhone besides Apple.
> Are you saying that their no competitors on iOS
None with the same competitive advantage as Apple. Whether that is fair or not seems to be up to individual markets.
So are you saying that
- Apple Music is more popular on iOS than Spotify?
- The iWork’s suite is more popular than MSOffice or GSuite?
- AppleTV is more popular than Netflix?
- Apple Books is more popular than Kindle?
Well, neither the hardware or software supports PCI access, so... connectivity would be a start. It's about time iPhone supported Thunderbolt anyways.
> So are you saying that {Apple Music is more popular on iOS than Spotify?, ...}
I'm saying that Apple exerts a literal anticompetitive advantage by choosing their own margins (among many other things). Nothing about legislation like the Digital Market Act accuses Apple of monopoly. It designates them as a Gatekeeper (a position Apple chose) and holds them subject to additional guidelines. That is the cost of owning a userbase in Europe, now.
How is Apple “choosing their own margins?” None of the example I listed pay Apple one penny for in app purchases.
I guess since the tech industry in Europe can’t export anything worthwhile, passing legislation is the best they have.
Let's see!
I, for one, would be very excited to watch Apple admit that the European audience isn't worth their while. After all, Apple acquires their companies year[0] after year[1] after year[2] after year[3] after year[4] after year[5] after year[6] after year[7] after year[8] after year[9] after year[10] after year[11] after year[12] after year[13].
Well, maybe Europe matters to Apple... b-but not much!
[0] https://en.wikipedia.org/wiki/BIS_Records
[1] https://www.forbes.com/sites/ronshevlin/2022/03/27/why-apple...
[2] https://en.wikipedia.org/wiki/Primephonic
[3] https://www.macrumors.com/2020/04/03/apple-acquires-ai-start...
[4] https://en.wikipedia.org/wiki/Intel_XMM_modems
[5] https://en.wikipedia.org/wiki/Dialog_Semiconductor
[6] https://en.wikipedia.org/wiki/Beddit
[7] https://www.businessinsider.com/apple-acquired-finnish-compa...
[8] https://techcrunch.com/2015/11/24/apple-faceshift/
[9] https://www.pcmag.com/news/report-apple-buys-digital-magazin...
[10] https://techcrunch.com/2013/08/28/apple-reportedly-acquires-...
[11] https://www.macrumors.com/2012/05/30/apple-acquiring-italian...
[12] https://venturebeat.com/mobile/apple-buys-c3-technologies-on...
[13] https://en.wikipedia.org/wiki/Polar_Rose_(facial_recognition...
...I don't even give a half-shit about Europe, either. I just find it funny how their sovereignty pisses off corporate sycophants so much. People like you actually get worked-up into writing comment-after-comment of non-argument, refusing to write something more substantiated than ad-hominem attacks. All I have to do to refute you is cite my sources.
Of course you would also need a power plug.
And this is why when you leave geeks to their own devices they end up designing something like the Homermobile.
And none of “Apple’s competitors” pay a penny for in app purchases to Apple. They all force you to subscribe outside of the App Store.
> you can’t use CUDA - which only works on Nvidia hardware
Tell me you've never used CUDA without telling me you've never used CUDA.
Before Apple revoked Nvidia's certificates, they in fact did support CUDA on Apple hardware. You needed an Nvidia GPU to run it, but Apple hardware and software is a perfectly usable client: https://www.nvidia.com/en-us/drivers/cuda/mac-driver-archive...
Failing that, Nvidia still maintains BSD-style drivers if Apple did let them install their drivers. Now you're going to yell at me about how this is an unrealistic workflow, but plugging your iPhone or Mac into a TV, DAC or external drive isn't.
> just in case someone wants to carry around an Nvidua breakout box to do intensive GPU work on an iPhone.
Sure. They exist, nothing on the technical side stops Apple from implementing it. My point is not that they must add the feature, but that it is totally possible if Apple didn't arbitrarily limit userspace drivers. Draw your own conclusions about it, I couldn't give less of a shit.
If changing the iPhone to Thunderbolt and supporting eGPUs in software is equivalent to a "Homermobile" in your head, get a grip.
> Now you're going to yell at me about how this is an unrealistic workflow, but plugging your iPhone or Mac into a TV, DAC or external drive isn't.
So how large do you think the market is for an external GPU on a phone? If there were a market for it, why isn’t anyone doing it on an Android phone?
We are talking about iPhones - not Macs. If you can engineer an Nvidia breakout box for ARM based Macs, there is nothing stopping you from turning off System Integrity Protection and installing an unsigned driver and you doing it yourself.
> Failing that, Nvidia still maintains BSD-style drivers if Apple did let them install their drivers.
You can turn off SIP right now and do it, good luck getting that working on an ARM based Mac where even Apple hasn’t architected a method yet to use third party GPUs.
> Sure. They exist, nothing on the technical side stops Apple from implementing it. My point is not that they must add the feature, but that it is totally possible if Apple didn't arbitrarily limit userspace drivers. Draw your own conclusions about it, I couldn't give less of a shit.
If there were a market for it on phones why doesn’t it exist for Android?
> If changing the iPhone to Thunderbolt and supporting eGPUs in software is equivalent to a "Homermobile" in your head, get a grip.
Yet no Android supports an external breakout box for eGPUs. Maybe it is a Homermobile?
The iPad Pros do in fact support Thunderbolt.
It is really funny watching you be butthurt about European regulators though. Can we go back to that, or will I have to wait for the bill to hit frontpage again? If you're done complaining about that stuff, I don't think we have much more to talk about (sorry for cutting out the eGPU detour, hope you had fun).
You could if you had the technical prowess create an Nvidia breakout box for the Mac right now Apple doesn’t stop you from doing anything you want. You brought up Apple stopping you from running CUDA on Apple hardware
Hopefully you realize the original Europe claims were absurd. But uhh... up to you pal. Clearly we live worlds apart.
> You could if you had the technical prowess create an Nvidia breakout box for the Mac right now
You can't even do it on Linux without official Nvidia drivers, which Apple refuses to sign for MacOS. Nouveau is veritably broken on modern GPU hardware; you either run the blob or get 2D raster output at 20FPS and no power management.
And what part of you can disable system integrity protection on MacOS and run unsigned drivers is hard to comprehend?
It's not hardware.
Would you be happier if I said "excluding software that competes with their own"?
Why would it be any better if they are excluding hardware that competes with their own?
> Apple never stopped VSCode from being on iOS. There are plenty of IDEs for the iPad built by third parties.
VSCode is an app that needs the .NET runtime, in order to run the code you write in e.g. C#. They don't allow the runtime it needs which means they don't allow the app (which competes with theirs).
> Are you making up things now? You don’t run CUDA on any non Nvidia hardware.
iPads have Thunderbolt, so you could actually plug NVIDIA hardware into them. Or someone could port CUDA to Apple's GPUs, as AMD essentially did with HIP. Except that would be this:
> But yes, you are free to port OpenCL to iOS if you want to.
Implementing such a thing efficiently requires driver support, but you can't modify the drivers nor supply your own.
> A “payment API” is not an “application”. Yes the cross platform frameworks do expose the underlying API.
A payments system is a part of an application. They prohibit broad categories of apps that use any but their own payments system.
Using Apple's payments API doesn't allow for apps that use a competing payments system, which was the whole point.
> And you can have a cross platform app that forces you to subscribe outside of the App Store - Spotify, Netflix, Microsoft, etc all do that.
I point to a category of thing they prohibit and then you point to the exception that proves the rule as if that meant they don't impose the rule.
We know they don't do it for a specific subset of subscription services anymore. They still do it in general, which is to say that they still do it.
> Are you saying that their no competitors on iOS to AppleTV, Apple Music, Apple Podcasts, Apple Books, the iWork suite?
This is the same thing again. It's not a question of if they prohibit every app that competes with theirs, it's a question of if they prohibit any app that competes with theirs.
Which they do. You can't get real Firefox on iOS:
https://support.mozilla.org/en-US/kb/add-ons-firefox-ios
And it's not because Mozilla doesn't want to provide it.
So now are kernel extensions also “applications”?
> VSCode is an app that needs the .NET runtime, in order to run the code you write in e.g. C#.
You could not possibly be more wrong. VSCode is written in Typescript. It is an Electron app. There have been cross platform JS frameworks that ran on iOS for a decade.
Besides that, it’s been years since you have needed the .Net runtime to run .Net applications.
https://learn.microsoft.com/en-us/windows/uwp/dotnet-native/
And Microsoft already has a solution to run .Net code on iOS
https://dotnet.microsoft.com/en-us/apps/mobile
> payments system is a part of an application. They prohibit broad categories of apps that use any but their own payments system.
There is an existence proof that there are applications that exist on iOS right now that force you to pay outside of their ecosystem.
Are you really crying over pay to win games?
> We know they don't do it for a specific subset of subscription services anymore. They still do it in general, which is to say that they still do it
Can you name any none pay to win games that couldn’t be a subscription service outside of the App Store?
And your solution to supporting CUDA is a breakout box on an iPad with an nvidua card? Who the hell is going to buy that?
There is nothing stopping
They're things some "applications" need in order to function, so if you can't include them you've effectively prohibited those applications.
But what does it matter if something is an "application" or not? Why is it any better if Apple precludes third parties from competing with their kernel extensions?
> VSCode is written in Typescript. It is an Electron app.
It's not the IDE you have to be able to run, it's the code you write in it.
> And Microsoft already has a solution to run .Net code on iOS
That's just compiling the code to a native binary, which you would then have to go submit through Apple's store. How does that help for an IDE expected to allow you to test (i.e. execute) and debug the code you've just written ten seconds ago?
> Can you name any none pay to win games that couldn’t be a subscription service outside of the App Store?
https://developer.apple.com/support/reader-apps/
We can see right there some examples of what isn't allowed:
Not facilitate real-time, person-to-person services (e.g., providing tutoring services, medical consultations, real estate tours, or fitness training).
Apps that let people access digital content such as music or video, but not as the primary functionality, are not considered reader apps and are not eligible for the External Link Account Entitlement. For example, a social networking app that lets people stream audiovisual content is not eligible.
> And your solution to supporting CUDA is a breakout box on an iPad with an nvidua card? Who the hell is going to buy that?
Presumably the same sort of people who buy the existing graphics card docks on the market for other platforms.
It can run anything that you can run on an x86 in user mode. I downloaded the AWS CLI (which requires Python) to run some tests
By the way, you were completely wrong about VSCode being written in .Net.
> That's just compiling the code to a native binary, which you would then have to go submit through Apple's store. How does that help for an IDE expected to allow you to test (i.e. execute) and debug the code you've just written ten seconds ago?
There is an existence proof that it could be done. If you ran iSH with remote VNC you could have a full IDE on a Mac.
> We can see right there some examples of what isn't allowed: Not facilitate real-time, person-to-person services (e.g., providing tutoring services, medical consultations, real estate tours, or fitness training).
You lack imagination. I actually know someone who does fitness classes online where most people use iPads and doesn’t pay Apple a dime - she’s sleeping right next to me. She uses Zoom.
> But what does it matter if something is an "application" or not? Why is it any better if Apple precludes third parties from competing with their kernel extensions?
It matters because once you allow kernel extensions it can bypass the security sandbox.
Do you also think that Apple should allow third parties access to the Secure Enclave?
Apple got asked why they take 30% by the US gov, answered that they just like money, and never got asked anything again. The only gov that could do squat on that front is China. Netherlands and Korea barely got some small exceptions.
Same way Google's only trial that went anywhere regarding Google Services binding happened in EU, and the only country that got away with a vibrant android ecosystem is China (to my knowledge...does India do too?)
I'll poopoo Eu's efforts when another entity actually splits Google away, and I'll be really happy to do so, trust me.
Chat control is not really tech regulation though, even if it affects tech the proposal is written by people who have no idea what the fuck they are doing. Ylva Johansson (The MEP responsible for it) has even said straight out she has no idea if or how this should be complied with but "They will figure it out" (They being tech companies). With people like that deciding our laws of course you end up with shit like this.
I would disagree with that. I like being able to download a copy of all my data from all these big websites. Since I no longer use Twitter/X, it's nice to have a copy of everything I have posted there.
Why would you want a law with high compliance costs to solve a problem that was already solved? It makes no sense to impose the costs unless the thing it purports to solve actually presents as a problem.
Product design doesn't need to be written into law.
Please stop repeating that argument, it's not true. There's periodic review proces built in into this legislation.
>attempted to enforce a similar mandate for micro-USB back in the day - can you imagine if they succeeded?
That would be great. Probably Apple would transition to the USB-C a few years ago with everyone instead current nonsense.
Lightning originated when there was a talk about mandating micro-USB and finally died, after 11 years, when USB-C was mandated. If not for that legislation it would still be alive.
It's actually faster that way.
Now it will be a chicken and egg problem. No one will want to put a new connector in phones, because they'd need usb c too anyways. So no new standard will emerge with wide enough industry support. If the micro usb "mandate" was as binding and restrictive as this one, we would not have had usb c.
This is why the USB Implimentation Forum exists. If you want to put a new connector in phones, do it; just do it through the proper, standardized and official channels. If every manufacturer had their own competing charging standard, we would be back to square 1 with 15 different ways to route serial data and 15w of juice. As companies like Apple have so gratefully demonstrated, most companies tend to invent those chargers for licensing royalties rather than a willingness to push the envelope on charging or data transfer technology.
And plus... it's a serial cable. The demand for "research and development" has remained the same since they figured out how to flatten the type B connector. You can't fool anybody by releasing a 'new and improved' cable or connector that is just the same serial bus with proprietary mumbo-jumbo on top.
And this sums up the issue perfectly. “I don’t want improvements. I want design by committee least common denominator”
I think I'll give EU a chance before I will agree that they created bottle neck by this legislation.
The EU just mandated a charging standard. It didn’t mandate that Apple go whole hog implementing the USB 3.1 protocol including video over USB C - all things that the typical Android phone doesn’t do.
When do you think mobile devices will be limited by USB-C? Another 20 years? Is it worth for you to allow mess, vendor locks and ripping of customers for another quarter of century, because some magic connector that solves problem we don't know exists can be delayed by 2-3 years? Really?
Anti-competitive practices from corporations delay progress way more than that.
If micro usb was enforced, USB C in its current form wouldn’t exist.
A periodic review process won't matter. Somebody has to actually research, design and create an alternative. Why would they spend resources on that if their competitors can just decide that they're not allowed to use it?
Not at all, I think it achieved exactly what it was designed to do. The part where we blindly accept cookie banners without holding companies accountable for doing so much tracking (or put them in place without questioning if they're needed) is totally on us.
GDPR did give us the tools and the means, we just don't take full advantage of them.
It's the main reason EEA countries show the full price of something on their price label, as well as unit prices (e.g. cost per kg) in most cases. You may not be aware of it if you're outside of europe though.
How is GDPR any good? All I saw is addition of misleading popups in the sites without any actual improvement to privacy. Has any big tech reduced data collection after GDPR?
Any big tech still operating in the EU did. A lot.
No doubt it made them spend a lot of money on lawyers to figure out the best way to keep doing it, but what good is that?
Pre-GDPR, you'd open your logs to any random company promising to come up with something that could remotely help your marketing department. Nowadays, you'll have to vet your partners and audit how your data is handled internally. It's hard to come up with concrete examples when it's basically every single company above a decent size, as they typically don't want to work against the law.
Same for the retention period, the very type of log that you'd open for internal harvesting (you could still be collecting the info, but keep it way more confidential to the point no one outside of your SRE team has access to the totality of it)
It's as if you were asking what changes the PSI and DSS compliance laws had on the card processing field.
And actual changes happens on many of these cases. An entity getting fined three of four times for different offenses doesn't mean they never fix any of the issues.
An easy example could be Google, who've been a target of so many of these complaints, and made many adjustment, up to building separate data centers with separate management rules to deal with the EU situation.