Moving on from interoperability in messaging services, there’s a lot more in the proposed Digital Markets Act. For instance, bloatware and other preinstalled applications on iOS and Android devices must be removable by the user, and users must be given choice of which browser, e-mail application, etc. they want to use by default. This is a complete no-brainer, and something virtually every user will welcome.
There’s also a lot of measures regarding data transparency and advertising. For instance, smaller companies that sell goods on e.g. Amazon must be given access to Amazon’s analytics and similar data. In a similar vein, people who buy ads on Google or Facebook must be able to assess the reach of their ads. And, of course, big technology companies will no longer be allowed to give preference to their own services and products.
These are all excellent steps in the right direction. Fines for violating the DMA will be massive – up to 10 percent of worldwide annual revenue, 5 percent of average daily turnover, and more.
Thom Holwerda,
It’s too early to celebrate, but assuming it passes then it’s about time! I applaud the EU’s efforts to reign in the abuses of our largest multinational corporations. It’s a shame that US representatives are too corrupt to represent consumer interests, but hopefully American SKUs can also benefit from these EU laws.
Alfman,
While there are some really nice changes in this group, I think they are including untested rules, which will eventually be harmful.
On the positive side, they are tackling force bundled apps. That is of course not that straightforward either (what happens if a user uninstalls a system service, making the device inoperable)?, it is overall a good benefit (looking at my phone with 10s of useless apps, many of which I cannot even disable).
That being said, messaging will be a thorny issue. I actually don’t care too much about “intellectual property” cries, but rather practicality of the user experience.
If I use, say WhatsApp, I have a certain expectation from the overall system. The party on the other end will be unable to record the conversation (at least not easily), and I can recall/delete messages sent to the wrong group. That will go away.
They will either have to do away with that functionality, or only have it on official clients.
(rest is speculation):
Then we reach the “imessage” problem with different colors: “green contacts” that use official client, and are secure, and “red contacts” that use 3rd party client that can potentially cause privacy issues. (Not that I like where Whatsapp wants to go wrt. privacy, fortunately they still could not change the TOC yet).
And then EU will force them to be the same color,
And then we will have years of court battles on these, with a time period of uncertainty.
And of course nation states will join as a side, not wanting to miss the opportunity to ban end-to-end crypto from civilian use.
And then it will probably end up with backdoors on all public messaging systems, just like SMS is today.
(end speculation)
I am not very optimistic. On one hand we need many of these changes. On the other hand, I am really pessimistic on where this will lead.
sukru,
Obviously not even whatsapp can guaranty such features will work once messages have been delivered. It’s a case of proprietary software using security through obscurity. You could technically make deletions part of the federated spec, but without the proprietary software, there’s no way to enforce it. Some delivery services deliver notices via standard email communications and then direct you to open the private message via a browser. It seems like this could achieve what you’re asking for and provide a delete feature that is as effective as whatsapp, no?
I think there are good solutions, but I don’t want to say too much here because I think it’s going to become redundant with the comments in the sister article that Thom posted specifically about whatsapp, messenger, imessage, etc.
Alfman,
You are correct, it is security by obscurity. However this is one of the rare cases where a DRM stack actually helps the user. If both ends have the same config, it is unlikely the app is compromised.
(Of course a bad actor can always take screenshots with a separate camera, even if you disable everything in the OS).
I have not seen the other article, yet. I’ll take a look, thanks.
sukru,
Just for the sake of argument there isn’t a technical reason you couldn’t include DRM as part of the federated network protocol. It could probably use the exact same DRM used by Hollywood (apple fairplay, google widevine, ms playready, etc). Or new purpose-made DRM with new features.
I’m sure you already know the problems of DRM, here are a few I can think of…
1) DRM depends on security by obscurity.
2) It often uses proprietary code that cannot be publically audited.
3) It hurts portability. Vendors may not care to port their DRM to a platform you want to use.
4) It blocks legitimate use cases like a screen reader to improve accessibility.
5) Years into the future DRM may render historic messages inaccessible to legitimate users. Consider old games that stop working in new operating systems due to unsupported legacy DRM.
I acknowledge that It’s important to discuss all the possibilities, but DRM hasn’t always had a good connotation with the public.
Alfman,
Sorry for stretching this out.
My argument was not DRM is a net benefit all the time. Just it had some positive effects for this specific use case (easy privacy barriers).
Anyway, all these are hypothetical. We will see the history as it unfolds very soon.
sukru,
Yes I understand. My point is that DRM still has the same cons even if we swap out Hollywood content with personal messages. Do the benefits outweigh the cons? It’s going to depend on who you ask 🙂
This sounds great in theory. In practice, I worry about unintended side-effects.
I don’t think it’s possible to define what an OS is (what it must include and what is “bundled on top”) in a strict/formal way suitable for lawyers and courts.
For example, for desktop; lots of Linux distros let you install and uninstall whatever GUI you like (Gnome, KDE, …), which proves that a GUI is not a necessary part of a desktop OS, and therefore the GUI bundled with Windows needs to be uninstallable?
For another example, for smartphone; iOS doesn’t use a virtual machine, which proves that a virtual machine isn’t a necessary part of a smartphone OS, and therefore the Dalvik/ART virtual machine bundled with Android needs to be uninstallable?
At first glance these examples seem extremely silly; but are they? Is Microsoft preventing fair competition with Gnome and KDE? Is Google preventing fair competition with Oracle’s Java virtual machine?
In a similar way, how do you define “an app”? If Microsoft compiled a web browser directly into the kernel can they say “Haha, it’s not an app, so you can’t uninstall it or change the default browser.”?
Perhaps the process should be that if a user can install some alternative software, set it as the preferred option and successfully reboot the computer to use the alternative software then and only then should the original software be uninstallable
nitram,
You just need to make sure the OS has a way to factory reset the original apps and settings at the owner’s request. It’s something that all operating systems should have IMHO such that no matter what userspace apps are broken it can automatically bootstrap the system by downloading and restoring the factory apps. To make it tamper resistant you could add it as a low level boot service.
The iphone already has several update & recovery modes so I don’t think it would be complex to support.
https://www.ultfone.com/iphone/comparison-between-iphone-dfu-mode-recovery-mode-safe-mode.html
If apple’s interested in paying me good money, I’d make it work 🙂
You are right, a factory reset should be available if the OS is completely broken.
However it would be nice if there was some mechanism to undo the last change if something goes wrong rather than losing all the modifications and settings that had previously been made and which had been working as desired.
nitram,
Yes, however that’s not what I meant. I was thinking a way to restore individual factory applications like the browser and app store without having to do a full factory reset.
It could be nice if apple implemented it’s macos “time machine” feature for IOS too.
If you use a checkpointing file system (like btrfs & zfs), then checkpoints are instant and rolling back is fast and easy but I’m not sure whether these filesystems/features are available for any mainstream mobile phones. Does anyone else know?
It’s going to be full of loopholes. Just like last time. lol
Personally, I am worried about this. The thing that motivates companies like Google and Valve to contribute a ton of open-source software is the business model of giving the OS away as open-source but making sure that the OS has a “preference” for the company’s online services by keeping some bits closed-source.
For example, does Google demanding that the Play Store be installed as part of GMS count as “preference”? Today, if you want Google location services, you need to comply with that “preference”. And you will comply, because even third-party apps like Citymapper require Google location services. If Google is forced to drop that “preference”, carriers can replace the Play Store in all devices they ship with all those “portals” they had before (things like Vodafone Live). And then there is Google Search, Gmail etc
So, does anyone believe Google will keep funding AOSP when they can no longer impose their “preference” for certain online services and that they will be basically giving code away to Amazon (and other Amazon derivatives worldwide) to make their own OSes? BTW I don’t think this law will change Android’s business model today, but it can for future products.
Lol, this could as well be the end of online advertising. With the exception of maybe YouTube apps, nobody pays attention to online ads, because they are essentially unregulated. So, people associate these ads with scams and view them with BS dectors set to “high”. For example, I associate ads with the “you are the 1.000.000th visitor!” scams and pre-assume they are scams.
Which is a problem considering how many services are funded by online ads. People nowadays think that buy buying a client computer and an internet connection, they are entitled to a variety of free services at the server side. For example, people think that image and video hosting is something that naturally comes for free.
kurkosdr,
Users could still have them, but manufacturers couldn’t be contractually forced to carry them. Companies like google and apple would have to rely less on their dominant positions and one sided contracts and compete on merit instead.
There are already questions about google’s commitment to android, specifically if google intends to displace it with Fuschia.
Anyways I think it’s a good thing if dominant companies are forced to compete on merit and less on monopoly benefits. Remember that bundling was always intended to give the dominant companies advantages over the free market. Let’s assume they loose these advantages and that they are no longer willing or able to be loss leaders after bundling gets broken up. Well that’s not automatically a bad thing. The market would have a chance at becoming more competitive and merit based.
I don’t believe ads are realistically going away any time soon, but if they did I think many people would consider it a positive outcome, so much the better! The economics of ads are very insulting to consumers; we ultimately have to pay more for products and services to subsidize the ads that we collectively hate so much. If we manage to turn back the clock on today’s incessant advertising, then I don’t think many people would mind.
http://www.waynesthisandthat.com/commerciallength.htm
Companies can still maintain API monopolies (up until Windows 7, Windows didn’t meaningfully force any service to you, save for IE defaults nobody cared about, but maintained a huge advantage via its monopoly on win32 and win64).
Also, notice the conspicuous absence of references to alternative app stores? This means the iOS model of having an API and App Store monopoly is not affected.
What this law could do is throw a wrench in the AOSP+GMS business model. If Google can’t monetize the GMS part, they have no business reason to keep paying for the AOSP part. Again, I don’t think this law will change Android’s business model today, but it can for future products.
People who don’t buy brand-name stuff don’t subsidize ads. They get a ton of free server-side services while the “ballers” who buy lots of stuff (and lots of branded stuff) subsidize those services. Imagine if you paid for every online service that involves video and images (aka expensive things to host) on the web. I wouldn’t consider it a positive outcome.
kurkosdr,
Yes, we always have to keep an eye on monopolies to make sure they’re not abusing their positions. I just want make something clear though: the problem isn’t so much when people legitimately chose windows. The problem is when they’re forced to buy or use something they didn’t need or want. Obviously these abuses are happening to alt-OS users all the time, but even windows users who own a retail copy of windows may be forced to buy a second copy of windows they don’t need. This is an example of abuse.
I disagree with the premise. P2P was already taking the role of delivering large content at scale before centralized providers ever could. It wasn’t scalability or financial problems that held these services back so much as a legal ones. P2P services were so successful in fact that RIAA and MPAA were terrified they couldn’t stop them. Industries were forced to bring competing services to market late to the game to fill the consumer demand. Centralized services have succeeded because they have large monopolies on legal content, but the reality is they had to spend tons of money to catch up to what the public could already do much more efficiency without them. So I don’t think we would be at a loss for video distribution technology in a world where centralized providers never existed.