If Apple had very quietly allowed sideloading a few years ago, that would have removed the antitrust threat – while the overwhelming majority of iPhone owners would have continued to get their apps from the App Store, just as they always have. The percentage of iPhone owners who will ever sideload an app is vanishingly small, so Apple has spent a lot of time and energy fighting a battle that is completely pointless
Worse, by fighting the issue so loudly and for so long, Apple has actually given the issue way more publicity than it would ever have received otherwise. It has turned what would otherwise have been a boring technical detail covered only by the Apple press into a mass-media news story. Apple has effectively contributed to its portrayal as a bad guy, with zero benefit to the company.
Right on the money – and I’m glad it has transpired this way. I doubt we’d ever have gotten this far without Apple continuously trying to stop it.
I also agree. Sideloading is quite a niche activity. It wasn’t necessary for apple to deny owners from having the right to install what they wanted for apple to still have a strong IOS app store monopoly. I do find it very problematic that competitors were totally blocked from the market for so many years. Even if the competitors only existed in the long tail, owners deserve to make their own app store choices without interference or coercion.
We all know why Apple locked sideloading on iOS:
1) Games like Fortnite have the brand power and customer loyalty power to encourage users to sideload the app (bypassing Apple’s tollbooth) in exchange for more in-game currency.
2) Apple doesn’t like emulators and how each of them comes with potentially thousands of games that can be acquired from outside the App Store
3) Apple doesn’t like Kodi either
4) …or browsers that don’t use the Safari engine.
5) You can’t downgrade apps either if there is no sideloading, so feature removals are final.
So, I disagree with the author. Lack of open sideloading in iOS was the reason Apple could keep an iron grip on the App Store. Instead, open sideloading on Android serves to keep Google honest, which is why the Play Store has things like emulators, Kodi, and browsers that aren’t just a reskin of the default browser. And yes, you can sideload Fortnite too.
So, I don’t agree with the author that the lack of open sideloading in Android is a minor issue. It was a major aspect of Apple’s control-freak-for-fun-and-profit strategy that just went poof. Good.
kurkosdr,
You are right about all the ways apple restricted the software owners could install & use, but I’m not sure the author would disagree with any of those points though. I for one agree with your points as well as the author’s points. I don’t see this contradicting his notion that sideloading is a niche activity for the masses. Have a look at android for comparison. While I am thankful to have the option to use alternative app stores, most people stick exclusively with google’s preinstalled store while everything else remains relatively niche. Do you have a reason to believe sideloading would be dramatically more popular with IOS users?
Well, as I read it, we’re all in agreement about apple’s authoritarian tendencies against owner control. But the author’s point was that the iron fist approach against owners may ultimately lead to worse repercussions for apple than letting a relatively small group of owners to do as they please with their own hardware would have done.
Alfman,
Another major issue is that Apple will not allow “competitor” apps on their marketplace. And they apply this retroactively.
A few years ago there was a sput between Apple and Spotify over app updates: https://www.vox.com/2016/6/30/12067578/spotify-apple-app-store-rejection . (Of course we need to listen to both sides. And, I can understand some of the reasoning, like providing a complete and consistent package).
If we had side-loading and/or alternative marketplaces, it would be a very different discussion, though.
sukru,
Indeed.
I sympathize with stores having the right to select and ban products from their own store shelves. If a store chooses not to carry certain products, IMHO it should be their prerogative to select the products they want in their store. But when one store has the means to literally block consumers from going to any other competing store via coercion or force, that’s clearly an abuse of power that we shouldn’t stand for.
I don’t believe courts should be dictating or micromanaging an app store’s wares, they should be free to set the terms for their own stores. However the book needs to get thrown at corporations depriving customers of the right to go use competing app stores. We should have put a stop to such overtly abusive practices at least a decade ago, but US antitrust regulators have a disappointing track record. For all I know, it might still amount to a mole hill.
Alfman,
And they already have understanding in this.
If I want the “Apple experience”, I can use the App Store on my Mac.
If I want the open source experience, I can use “macports” or “brew”.
And if I want something custom, I can always download stuff from the web.
But of course, they are also trying to limit those as well. For example, try running a kext for a driver:
https://github.com/macos-fuse-t/fuse-t
sukru,
Yeah, the same is true of windows. It is discouraging for FOSS developers on mainstream operating systems. I gave up developing my own windows drivers, which is something I did back in the day. Honestly I felt the windows kernel ABI was better designed than linux, the unstable ABI was breaking all the time. But microsoft’s usurping control over kernel policies from owners and developers was unbearable for me. That gave me the kick to switch to linux as they were far more welcoming of indy development. From that point forward everything would get implemented on linux first and windows became an afterthought for me as a developer. In retrospect, allowing FOSS developer relations to sour was a huge strategic mistake for microsoft culminating in the loss of multi-billion dollar hosting and HPC industries to competitors. But their leadership didn’t know any better at the time.
@alfman
Out of sheer curiosity, what is this hardware that needs kernel-space drivers? Microsoft is trying to limit kernel-space drivers in Windows and trying to push hardware vendors towards using generic interfaces or making the hardware appear as a serial port whose data are decoded/encoded by a user-space app.
And as a Windows user, I sympathise. The over-reliance on third-party kernel-space drivers forces Microsoft to keep ABIs stable for as long as possible, limiting innovation in the kernel. In a perfect world, the only kernel-space drivers would be for GPUs (which are generally well-supported by the GPU vendor with new drivers), and everything else would either use standardised interfaces (think AC97, HDA, or USB mass storage) which already have first-party drivers in the kernel or be a serial port whose data are decoded/encoded by a user-space app.
kurkosdr,
I was working on raid file systems. Speaking in terms of hardware though, I’d say this depends. For most USB peripherals userspace drivers are appropriate IMHO and I agree with you. But for things like video capture cards that can add more CPU overhead and latency. I had already switched to linux when I started working with realtime applications, but there too I’d say kernel space can be better.
@alfman
Why can’t a video capture card present a standard interface like webcams do? And why does a RAID controller need a driver? FakeRAID (aka software-implemented RAID) shouldn’t exist anymore.
This whole overreliance on drivers was warranted in the Windows 95 and Windows NT 4.0 era (when offloading some processing to the CPU made sense), but now it’s not warranted. Chips are so cheap that hardware can do the processing on its side and present a standard interface to the OS, instead of forcing a soon-to-be-unmaintained driver on the user that can also compromise system stability. MacOS did the right thing by curtailing third-party drivers and forcing hardware vendors to standardised interfaces.
BTW about the unsigned drivers thing, have you tried the following method of loading unsigned drivers?
https://www.youtube.com/watch?v=0s6IoQXBJNo
Same for that old scanner of yours in a previous discussion. You probably have, just putting it out there.
kurkosdr,
Maybe they should, but alas it hasn’t been the case for any of the capture equipment I’ve owned. All of them succumbed to driver incompatibilities over time
Real hardware raid has real benefits: less system overhead, more redundancy, etc. But I agree with you that those “fake raids” were largely pointless,
Note that the raid file system I was working on was not hardware based. Even with the benefit of hindsight I still believe a kernel driver was the best technical approach. A userspace implementation creates a lot more syscall overhead not only between processes using the file system, but to implement the raid itself, which is quite undesirable. Even on linux where FUSE file systems are easy to create, there’s a performance cost. Fundamentally this goes back to the debate between macrokernels versus microkernels, obviously they have pros and cons but we’d need a lot of ground to cover this topic, haha.
Yes test signing is one of the first to try, but I had tried it originally and I retried it during our discussion to no avail. A long time ago (I think it was vista) I had found two signed tools to load your own 64bit drivers and override microsoft’s policy restrictions, these worked temporarily but every update microsoft would block them again.
That he was able to get 3dfx to work is very interesting, I’m not sure whether glide needed a kernel space driver at all? 3dfx was kind of particular in that the hardware wasn’t handled by VBIOS/GDI/etc. The hardware was never shared and one process would have full exclusive control over it. There was a physical relay that bypassed the GPU and OS graphics stack entirely. That makes me wonder if the lack of OS integration could explain why it still worked after all this time.
So it’s software-RAID that works on any PC? Isn’t there some “standard” way to do RAID software emulation in Windows (for example using some software from some big company) other than using some driver someone wrote?
kurkosdr,
File system level raid isn’t quite the same as block level raid. The later was common, but not very elegant for thin volumes, file level snapshotting, file level corruption detection, etc. I was also interested in adding distributed features, though I didn’t get that far.
Today a ZFS, windows port could do most of what I wanted, but back then windows file systems were lacking (at least with the non-enterprise editions of windows I had). In principal I think owners should be allowed to write their own drivers without having to justify it to an upstream vendor. A sensible rebuttal might be that writing drivers is a very niche activity and it’s not a vendor’s responsibility to cater to us, but it’s this very attitude that makes me glad I left windows – developer needs aren’t sufficiently prioritized. Despite some rough edges, I do think linux is the better choice for developers like me.
Sideloading is a niche in Android because the Play Store has most things people need. There are emulators, Kodi, lots of browsers, lots of “competitor” apps etc. Google learned their lesson when they banned psx4droid and everyone was sideloading it.
Also, I don’t know how niche it is right now, considering you need to sideload the Epic Store to play Fortnite.
Google Play has emulators and browsers mostly because Google doesn’t care about what tech you use to build your app the way Apple does. Secondarily, they don’t charge nearly as much to list your app on the app store, which is MUCH friendlier to gray hat and open source app producers.
In short, its policy that allows that stuff on Play, and a little bit economic incentives – but the “threat” of side loading? Literally doesn’t matter. That’s not in the consideration matrix. They actively encouraged side loading entire app stores early on. They don’t care what software you run on your device.
They DO care about large partners/competitors forking Android, and removing Google Play and Google’s other platform support, but that’s a different thing entirely, and has nothing to do with side loading.
I’ve used custom Android builds for years in the past but never sideloaded apps on my phone — with the only exception being build of Google Camera that enabled support for my 1) custom ROM 2) officially unsupported phone. At the time Google Camera started rolling out the blurred background portrait effect and maybe also HDR+ support was limited to only a few phones / apps.
I don’t agree with such sentiment as the abuse from Apple side was premeditated and done on purpose. To force Apple into supporting normal installation of applications. Such effort took years. Now when Apple must support it and has no real say in it anymore. Saying it never really mattered for Apple in the first place. That just shows the level of ignorance and lack of understanding of the subject involved. This changes Apple in their core. But i understand that if you have being taking the blue pill for too long. It will take a while to snap out of it. Currently i see a lot of Apple iOS users complaining but now we will be “unsafe”. This is just crazy.
This applies to many categories of corporate group think, especially Digital Rights Management software and rootkits, and lawsuits against piracy. The thing the spreadsheet brains never seem to grasp is that most of the folks downloading stuff where not in the market in the first place – they were never going to spend the money, and it is therefor not a lost revenue (and piracy doesn’t cost the producer anything, because copies are free to make). So it makes no sense to spend millions trying to prevent piracy, when it doesn’t cost you anything (and EVERY pro-DRM “study” or model starts with the *assumption* of lost revenue, with no evidence, or even an attempt at evidence – just assertions. Most empirical studies show the inverse.) It’s the same thing in all the cases, DRM, or blocking side loading to solve whatever problem Apple says it solves (it almost certainly doesn’t), the solution is worse than the (expressed) problem.
CaptainN–,
That is not entirely true,
Music market has demonstrated people would stop pirating if there is a good alternative (drm-free mp3/aac downloads at reasonable prices), or good catalog with high quality streaming (spotify and 100 others).
Pirating has costs (low quality rips, bait-and-switch, malware, “dj” edits, etc). And those who have the means, could prefer to pay instead.
However there is a limit. If you charge too high, or don’t have enough content, people will go back to pirating. I think this is already happening in the TV land (too many services to subscribe to, and too many limits, no 4K, no offline, no sharing, etc).
sukru,
Surely it’s a function of both cost and convenience, And there are some who were never going to be paying customers. Investing in DRM to convert these non-customers is arguably of questionable value considering that DRM often hurts legitimate owners more than it stops the pirates. For example there are many times DRM software won’t work under wine. I know opinions will vary, but my point is that the tradeoffs are muddy and complicated.
Alfman,
Any purchase with a DRM is essentially a rental. So, yes, that should also be taken part of value considerations.
Probably, something like:
value = price x w1 + convenience x w2 + longevity x w3 + risk x w4 + quality x w5 + …
And the weights are dependent upon the user. For a student price would have a high negative coefficient, but for someone with a high disposable income, it could even be zero.
(And drm would be part of longevity in this example. In in your case, depending on platform it would also be part of convenience — do you want to dual boot to Windows just to watch Netflix?).