Hacker News new | past | comments | ask | show | jobs | submit login
Apple Accidentally Unpatches Vulnerability, Leading to New iOS 12.4 Jailbreak (macrumors.com)
134 points by dvcrn 62 days ago | hide | past | web | favorite | 172 comments



It's unfortunate how Apple and Google approach device ownership, and their attitude towards the concept of general computing is concerning.

We do not control our own devices, we cannot stop certain processes on them, and we do not know where our personal data is sent.

We either have to flash ROMs from questionable sources and apply temporary exploits to get some kind of resemblance of control of our own devices, or we have to spend years to learn the skills to unlock these systems ourselves.


I think Apple's approach is the only reasonable one for the general population. The technological complexity of any smartphone is far beyond comprehension for most people. I write iOS software for a living, and even with complete access to the source code, I couldn't reasonably evaluate my iPhone's software - let alone the hardware.

The idea that ROMs from questionable sources make your device safer sounds very strange to me.

Basically every electronic device has countless security issues. Some of them are found of which some are published of which most are eventually fixed (by rather large teams of professionals). In that regard, Apple could and should do better.

But the burden of making such a complex device secure simply can't be put on the end user.

While I would welcome deeper access for technically inclined people, I'm not sure that option can really be given by Apple/Google without the risk of becoming a disadvantage for many users.


I think locking down a system by default, but offering a way to gain elevated priviledges, while educating and properly warning users before certain actions is better than taking away everyone's control over their own devices, and therefore restricting their freedom.


The problem with that approach is $popular_social_media app comes along and coaxes users to relax said privileges "because reasons" and before long there's a signigficant proportion of users who altered the security model of their device without understanding what is going on.


Personal freedoms have always had risks, but is it really warranted to take them away in this case and not offer a way to get them back in any shape or form?

Looking at recent Samsung devices, is Google Services Framework really that integral to the security of my device that I must be forbidden from disabling that package? Isn't there an alternative way to achieve a comparable level of security, but without slurping up my personal data?

The consequence of security does not have to be a complete loss of control, nor the inability to prevent in a practical way the collection of our personal data.

Not to mention the whole security argument falls apart when perfectly fine Android devices are left without security updates 2-3 years after purchase.


I think there could be an argument that the personal freedom you mention, when risks are realized, can degrade the experience of the world at large. Lazy example: a botnet running on many machines compromised as described above sending spam email to innocents.

I’m still on the fence about whether that justifies their protocols. I think I actually lean toward “no”, but I’ve also lately become keenly aware of the difficulty of even simple things like keeping everything up to date, and my lack of real insight into what those updates include. If I’m effectively trusting them anyway, might as well trust them to get it to me ASAP, right?

I’m also enough of a realist to assume there’s a Fight Club style “A times B times C > X” reputational/financial risk logic going on here. If there’s few enough of the devices out there, it’s probably cheaper to apologize (legally, as in settle).


Apple is not taking away personal freedoms.

They create products, which they offer for sale. People can freely choose to purchase those products or not.

Apple has opinions about how to make products, which are embodied in the products they offer for sale. Other companies have different opinions. This is how a market is supposed to work.

I don’t think we should talk about Apple as if they are a government taking away freedoms. It confers too much authority upon Apple and too much victimhood upon customers.


That argument kind of breaks down when you have a monopoly on the market like Apple and Google do. If there isn't a viable option to their product then i'm not free to choose.

We shouldn't be talking about Apple like a government, but the government should probably be regulating this a bit better;


Are you seriously making the case we should be thankful about the restrictions they put on their products, because at least they exist? Is it somehow a problem to wish for things to be better, even though they could be worse?

They patently did contribute to the erosion of freedoms that people fundamentally used to have with their hardware.


> The problem with that approach is $popular_social_media app comes along and coaxes users to relax said privileges "because reasons" and before long there's a signigficant proportion of users who altered the security model of their device without understanding what is going on.

And the problem with a single signature authority, as we're seeing in China, is being murdered by an illicit state, or undergoing active discrimination in many other ways even in most western countries.


If people do that, that's on them. So long as the device appropriately warns people, I fail to see how it's the companies problem to baby people who don't know what they're doing. It's their device, if they want to break out, let them.

It's like saying "Why should we have knives? It's only a matter of time until $popular_social_media comes along and tells people to cut off their index fingers and before long there's a significant proportion of users who can't point anymore".


No, it's not at all like that. Knife users will be perfectly capable of understanding what they're doing when they cut off their fingers, and assessing its impact on them. They don't even need pages of warning text to explain this.

That's not the case for smartphone (etc) users.


A lot of tools have safety measures which can't be circumvented by their users (e.g., you have to use both hands to start thems). The reason being that some dangers are easily underestimated, even by experienced users. Manufacturers do indeed much better about the inherent risks of their products than users.

If a knive could be built which allows to cut food, and protects you from cutting off your index finger, wouldn't that be great?


Are you really comparing an object which holds a threat of blood loss, loss of organs and possible death as requiring similar safeguards as a phone?


"Appropriately warns people" is nearly impossible and shouldn't be brushed away as a non issue.


I think this is a valid concern, and perhaps verifiable. How many Windows user actually create a non-admin account to use for their everyday work? I find, anecdotally, that a helluva lot of them don't; in fact, the very idea is foreign to them.


I'd argue this is largely a question of defaults and ergonomics.

Most users will leave the default settings if they don't have an active need to change them. Easily usable (and understandable) tools and interfaces prevent most needs from arising in the first place.

Concrete example: The root account on many Linux distros is disabled by default. I've never felt the need to enable it, because sudo does everything I need. Secure default, useful tools, unlockable system.

Historically we haven't had either of those things. Poor design and implementation led to bad choices by clueless users. The resulting mess is used as an excuse to restrict freedoms. The cure is arguably worse than the poison.


I agree with you. However, my point is that these people don't see the need to create the non-privileged account, even though there is a strong one.


Ok, popular app requires elevated privileges, gets compromised, removes any remote control Apple might have to stop/cleanup the mess, and then breaks iOS

Now what? Millions of users have to wipe and restore their phones or throw them away and buy new ones because someone's app trashed the phone? That would cripple Apple


Realistically what would you gain by doing that?


> The idea that ROMs from questionable sources make your device safer sounds very strange to me.

It's about owning your hardware, not safety. A person that's willing to go through the hassle knows the consequences of such actions and how to deal with them.

Do "normal people" need to do that? Absolutely not. Should it be easy to do that? Absolutely not. But for those of us that really want to own our hardware, there should be a way of doing so without relying on exploits.


I totally agree, but I don't have any good ideas on how to implement that. I'm not even sure if such a barrier should be technological or legal.


Yeah, pointing to the problem is quite easier than figuring out a solution.

As far as I'm concerned, I actually read the popups and having to click okay about five or six times in a row would make me second-guess my decision. Would that work for everyone? Most of the people? Some of the people? I don't know that answer.

> I'm not even sure if such a barrier should be technological or legal.

My answer would be both. I highly doubt it's in hardware manufacturer's interest to figure out a technological solution, but if there's some legal incentive for them to at least try, they'll figure out a technological solution.


Android's option of connecting the device to a computer over USB, running a program on the computer, logging into the device, seeing a scary warning and wiping the device seems to work well. I don't think I've heard of a large number of people being tricked into unlocking their bootloader — at worst, a handful of script kiddies might have been tricked.


Well, they didn't say the ROMs made their device safer, just that they gave them some semblance of control.


> The idea that ROMs from questionable sources make your device safer sounds very strange to me.

The first step on Android is usually to unlock the device boot loader in order to flash a recovery that will allow to erase partitions and install a tarball of the system. I saw no tutorial suggesting to re-lock the device boot and I bet people rarely do it.

This means anyone can take the device, boot it into recovery, plug it into USB and throw some adb/fastboot commands to do anything they want. Device encryption becomes moot because neither the recovery nor the bootloader can be trusted.


You don't relock the bootloader because every lock/unlock cycle clears user data, and you might need to update recovery to update your ROM. Even if you relock the recovery, next step is the infamous "no sha1 signature found, flashing boot sector unconditionally" (which is a step up from md5!) TWRP has enough attack surface; android devices have very little physical security.


You don't relock the bootloader because every lock/unlock cycle clears user data, and you might need to update recovery to update your ROM. Even if you relock the recovery, next step is the infamous "no sha1 signature found, flashing boot sector unconditionally" (which is a step up from md5!)


> I think Apple's approach is the only reasonable one for the general population.

The general population uses macOS just fine.


>The technological complexity of any smartphone is far beyond comprehension for most people.

The exact same goes for regular computers and operating systems. And we still have those. Shouldn't we at least have the option on mobile?


> And we still have those.

You have computer form factors and OS'es invented in the 80's and 90s for desktop. The only reason they still work that way is because of legacy/inertia.

The exception is the ChromeBook and CromeOS and that works similar to a smartphone. For a reason.


Chrome OS now has Crostini, which allows you to install basically any Linux app. Or is that also attributed to the legacy/inertia of Linux?


> even with complete access to the source code, I couldn't reasonably evaluate my iPhone's software - let alone the hardware

That's not the point. The point is that the community will be able to do it and auditors' lives will be much easier, which benefits everyone because it vastly increases the likelihood of an issue being found.

> The idea that ROMs from questionable sources make your device safer sounds very strange to me.

On Android by default most vendors ship a lot of bloatware and have demonstrated almost infinite incompetence or malice of both. While the developers "custom ROMs from questionable sources" (XDA forum threads) may not be experts in their fields at all and are quite likely to misconfigure the software possible creating some new holes, at least those images are compiled from open sources and are not the terrible manufacturer OS (I wouldn't onlike bank on stock Xiaomi software).


> The technological complexity of any smartphone is far beyond comprehension for most people.

Don't use it then. By shielding the stupid we are creating more stupid.


That idea is stupid in itself.

The whole of civilization has been a process of shielding people from having to know stuff.

The same way you don't know how to make fire from first principles, fix your car, make a CPU, or whatever...

Even someone with a Ph.D in computer hardware is shielded from tons of complexities and never has to know the whole process end to end.


But they are allowed to learn, which Apple doesn't want you to do. They could make it so your mom doesn't get root by accident, but you would still have the right to do so.


> They could make it so your mom doesn't get root by accident, but you would still have the right to do so.

How? Serious, genuine question. How can they give you “the right to do so”, but prevent “mom” from accidentally doing so or worse, having someone do it to their phone without them knowing?


I don’t think anyone is suggesting something similar to “This app would like to access your camera” but for unmitigated root access.

It could even be as involved as getting an official image from apple/google that allows root access. I don’t think a “mom” would accidentally download an image and flash their telephones.


> I don’t think a “mom” would accidentally download an image and flash their telephones.

Who says “mom” flashed her phone?

How does “mom” know the shady place they took their phone to repair a broken screen didn’t do it? (P.S. many documented cases of this)

How does “mom” know their jealous/cheating/whatever spouse/bf/gf/whatever didn’t do it? (P.S. many documented cases of this)

How does “mom” know the phone they bought off someone didn’t do this? (P.S. many documented cases of this)

How does “mom” know their kid didn’t do this so they could install stolen games? (P.S. many documented cases of this)

The list goes on and on.


These concerns all apply to existing jail breaking methods. An official image would be much easier to make obvious that it’s not the same as the non-rooted image.

In order for any of those threats to work it would require physical access and access to passwords and accounts (for example reconnecting to cloud services, restoring backups or even just unlocking the phone to perform a flash). At that point I don’t think it’s having or not having root access that is the issue.


> These concerns all apply to existing jail breaking methods.

Of course they do, hence my comment.

> At that point I don’t think it’s having or not having root access that is the issue.

That’s the most naive thing I’ve seen said all week.


Android's bootloader unlock process prevents mom from doing this just fine.


I have a food processor that won't run unless fully assembled and locked. Is that anti consumer?

My coffee grinder won't turn on without the lid locked in place. Is that anti consumer?


Goodness me it gets tiring seeing a post like this every time an article about iOS comes up.

It's a legitimate point to make, but it should rather be that we talk about the content of the article rather than bring out the 'Apple's taking our freedom away' soapbox.

It inevitably turns into a debate between one side who values personal freedoms but won't be told to buy Android phones widespread safety, security, and privacy and tell the others to just buy Android phones.

This goes around and around, usually getting nobody anywhere, and happens all over again the next time an article about iOS turns up on the front page.


Well there aren't any great alternatives. The state of general computing (as parent puts it) is frustrating, and people talk about it, sometimes by switching from a related topic. It's venting. Are there more productive ways of dealing with this? Maybe; but seemingly part of the frustration is that the average consumer feels powerless. Even your meta-post could be classed as a way to deal with this frustration. And, I wouldn't be so quick to dismiss venting as a waste of energy: it raises awareness, and in the right setting it can lead to discussions of the cause, and what could be done about it.


but seemingly part of the frustration is that the average consumer feels powerless

The average consumer could care less. They want to make phone calls, send text messages (over iMessage), surf Facebook, and take back to school pictures of their kids - and they just want it to work.


We need something similar to Godwin's law where anyone who argues that people don't care about something automatically loses the argument.

Of course people don't care more about iPhones than their own kids. Pick just about any X and people don't care more about X than their own kids.

That doesn't mean X doesn't matter, or people wouldn't care more about it if the understood it better, or wouldn't be better off if it was different, or that we shouldn't try to do something about it.


If we polled all smartphone owners and asked them how much they would pay to be able to own their phone in the way that Stallman owns his computer I'd bet a large amount of money that the median would be 0 cents and the average would be less than 1 cent. This isn't just a question of caring less. The huge majority of users do not care at all about this.


If we polled all smartphone owners and asked them who Stallman is, the majority of them wouldn't be able to tell you. That doesn't mean it doesn't affect them.

And if you're talking about a poll where you ask people whether they would like to have a larger selection of apps for their phone, or they think it should be easier for small developers to enter the market, or they think Google should have more control over their phone than they do, the majority of people are not going to choose less apps and less freedom at the same price.

The fact that you have to put it against dollars to get them to do what you want is just leveraging the fact that the median user is not rich to claim that nobody wants something that everybody wants because everybody also wants to save money (and the implication that the median person values it at less than a penny is hyperbole). Why should you have to pay extra for freedom? Should only the rich have it?


What you're really getting at by suggesting we poll people on Stallman, I think, is that they're uninformed, and AnthonyMouse is dead-on where he suggests rephrasing what it means to own your high-tech stuff so that it's comprehensible to the general public.


> The average consumer could care less.

You're presuming that the average consumer is uninformed, which I can understand, but it's not true in my experience. Most people I know are well aware that they're in a pickle.


So let me ask, why make this comment? Why reply like this? To shut people's concerns down? There's no value here. There's nothing to be gained. Conversely, sometimes it's OK just to let people complain: without the sentiment, how do you know if people still care?


It’s pretty obvious what the point is. It’s to complain about the soapbox / complaints in hopes of seeing less such comments in the future.


It's off topic and people should start getting banned for it because it turns the comments into useless bikeshedding


It's an article about enabling a jailbreak and you are mad that people talk about jailbreaking needing to be a thing?


If a problem is presented, someone is going to bring up a solution. The right solution has been the same for a long time, Apple hasn't implemented it yet. Maybe they'll never implement it, but maybe someone will consider alternatives because of the comment.


Agreed. It gets very tiring. Every post related to Apple or Google devolves into the same exact conversations, regardless of the content of the article.


> This goes around and around, usually getting nobody anywhere, and happens all over again the next time

A security regression in a new software release is a much older type of story that just gets talked in circles too.

>it should rather be that we talk about the content of the article

iPhones are jailbreakable again. Whoop-de-doo. What are we supposed to say about it?

Might as well talk about the system design that makes owners of devices getting root on them considered a flaw in the first place.


You can get a developer account, build and sign your own executables, and run whatever you want on your iPhone.

I don’t think those flashed ROMs give you appreciably more “control” over your iPhone than stock iOS provides, because actual control requires usable control surfaces. More likely, you are replacing the control surfaces provided by an accountable entity (Apple) who has prioritized your security and privacy and provides a constant stream of updates to maintain that, with what exactly?

The entire iOS feature set is designed to protect your personal data, from outside attackers who would seek to compromise it, to insider threats like apps trying to siphon off more than you might expect, to end-users inadvertently giving away their own (or your) data without a care in the world.

The security that a modern iPhone provides to its owner is truly a remarkable and commendable experience overall. I am extremely happy we as consumers have the choice to purchase exactly such a device.


The iOS feature set is intended to protect Apple's business model and revenue streams by forcing you to do things in Apple's jail. Getting a developer account is obviously an absurd and impractical approach to distribute software and proposing this is frankly ridiculous. It's not free, requires you own a Mac, and is way beyond the technical abilities of most users.


> requires you own a Mac

I mostly agree with you, but you don't necessarily need a Mac in order to use your developer account for sideloading—you can also use Cydia Impactor which is available for Windows/Linux.


> You can get a developer account, build and sign your own executables, and run whatever you want on your iPhone.

Sure, for an extra $100 every single year. Apple charging money for a feature does not protect users.


> Apple charging money for a feature does not protect users.

The $99 is not likely to make them money. It is a token fee to protect the app store from the simplest spam and scam apps.


Sure, it's a mere token for Apple, but it is a significant, recurring financial and logistical hurdle for me when I simply want the ability to use my pocket computing/surveillance device -- which I've already paid for -- in ways I deem fit.


This is a pretty significant shifting of the goalposts.

The argument was users should have the ability to run whatever code they want on their iPhones. That is actually possible today.

It costs a $100/year, which considering the costs of the phones is pretty reasonable. Part of the reason it has to cost something non-trivial is because otherwise it would encourage massive piracy, which would devalue the entire App Store (in fact exactly what we see on Android).

If the argument is users should be able to run whatever code they want, but they also must have free access to the development tools and resources like Xcode and Dev Center (which cost how many tens or hundreds of millions to develop?) then you’ve totally lost me.

I understand and appreciate the principle that it should generally be possible to develop and run the programs of your choosing on smartphone-type hardware. In no way should a company be forced to spends millions of dollars to facilitate that at scale if that’s not their business model, particularly when it would primarily be used to directly attack their ecosystem.


> The argument was users should have the ability to run whatever code they want on their iPhones. That is actually possible today.

Technically possible, fine. But there's a big gulf between "possible" and "not extremely painful".

Let's say I create a personal fork of the open source Bitwarden password manager, to add some trivial quirk that makes the software better fit my life. How do you propose I actually use my custom version without paying Apple an extra $100 per year?

Every 7 days, my version of the app will suddenly refuse to launch, until I get back to my computer and re-sign it. I would need to create a weekly calendar reminder, and never go on vacation without a computer nearby. Oh, and I'd better not have more than three of these forks, because that's another limitation for free accounts.

The 7 day limit is not Apple refusing to provide "free access to the development tools", it's an artificial restriction explicitly created to make running un-blessed code impractical for more than rudimentary testing.


It sounds like the problem you have isn't that it's not possible, but that it's not free (as in beer).

The free account is good for playing around with the environment and learning how to write code for iOS. It is not well suited for running production software on your phone. If you want to run production software on your phone, blessed or unblessed, you probably want to pay Apple $100/yr for the longer duration and higher app limit.

In my opinion, if you could do what you wanted for free, it would contribute massively to app piracy, and devalue the work of millions of developers on the App Store. $100/year is at about the right level to dissuade most people from circumventing the App Store (the average Apple user spends ~$75/year on the App Store).

> ...it's an artificial restriction explicitly created to make running un-blessed code impractical for more than rudimentary testing.

Crucially, paying Apple $100/year does not mean they ever see or have to bless code you deploy to your own devices. It just removes the limits in the development environment!


Allowing side-loading would probably contribute to app piracy somewhat, but I don't think it would have as large an effect as you describe. Android's customer base is less wealthy overall, and less likely to have payment information in the play store. I would look to Mac rather than Android for a more realistic picture of what the piracy landscape would look like. It's not bad.

But, I also fundamentally believe we shouldn't be restricting user freedoms to protect copyright.

$100 a year is completely and utterly cost-prohibitive for a lot of people, even within the context of someone who already owns an iPhone. A lot of children begin coding so they can create something for themselves, or change one thing in an application. With a 7-day limit, who would want to do that? While it's true that free accounts can technically experiment with coding, there's not much incentive when you can't really _use_ anything you've created.

I worry about a generation of children who is given iPads rather than computers. Sure, most of them would never have touched code anyway, and that's fine. But iOS completely removes the incentive to learn and explore—to actually hack and tweak and create the tools we use everyday, instead of blindly consuming them.


> I worry about a generation of children who is given iPads rather than computers.

I’m pretty sure they said the same thing about shells when GUI came along.

> But iOS completely removes the incentive to learn and explore...

I guess I just have an entirely different perspective on this. If the PC was a bicycle for the mind, what we have now is a veritable rocket ship. Consider the devices, peripherals, platforms, APIs, connectivity, distribution, and tooling that is available today for anyone with any interest in creative artistic expression, be it coding, non-textual programming, or otherwise...

Anyone with a inventive flair is going to look at the mind-bogglingly advanced technology that a billion people are carrying around and just salivate at the opportunity that provides.

Modern devices and the modern Internet may placate the masses, but they are likewise catnip and catapult for anyone who wants to code the next great solution for Problem X.


If you want to develop applications (or run arbitrary code) on their stack, which they spend billions of dollars developing and maintaining, yes, it will cost you $100/year.

Or you can get a free dev account, but the feature set it more limited and signatures are only good for 7 days.

Apple charges money for features so that they remain in business to keep making more features, and security updates too. So it is, in fact, exactly how they protect users.


> And signatures are only good for 7 days.

And again, how does that restriction protect users? As I see it, it's entirely user-hostile: it ensures any self-created apps aren't really usable.

If the limit was significantly longer, I would mostly shut up about all of this.


Who is this hypothetical person who can afford a Mac and the time to learn how to how to develop iOS applications and yet cannot afford $100 a year? Usually I'm all in favour of freedom, but I just can't see who's actually affected by this.


Okay, here's a personal example:

Apple does not allow dictionary apps on the App Store which actually use the word definitions built into iOS—they are required to provide their own definitions, which either take up precious storage space or are not available offline.

So, I found some old WTFPL-licensed code on Github, spent an hour or so futzing around to make it compile and look pretty on iOS 12 (because I had no clue what I was doing), and came up with this: https://github.com/Wowfunhappy/Dictionary. It works super well, and I use it every day on my phone.

The only reason I can use this app is because I'm Jailbroken.


You spent the time to Jailbreak the phone, but couldn't you also have just paid $100/year for the license to be able to build your own personal apps which can be deployed on your personal devices for 1 year at a time?


Well, that's not why I Jailbroke my phone. I Jailbroke my phone so I could:

• Install a Userscript to de-AMP pages in Google search results.

• Prevent Apple News from saving a history of what articles I read, thus disabling their recommendation engine and preventing a filter bubble.

• Add an extra row of app icons to my homescreen, so I can fit all my apps on one page.

• Get a warning when I set an alarm for PM rather than AM.

...and countless other little things.

Separately, I consider $100 an awful lot of money, especially for a subscription, which I try really hard to keep out of my life.


It's the person who doesn't know how to do that yet. Most people don't get started in software development by writing their own operating system kernel. They're using an existing application with published source code and want to make a little change. It's an ugly hack written by a teenager, they just want to use it for themselves. They'll get better at it as they do it more.

But now you say they have to buy a Mac and pay $100/year. Well, that's a no for that little initial change, so now they never get started to begin with.


Really? GNU/Linux doesn't thrive on billions and provides a near perfect functional alternative. Darwin/XNU is libre software, so it's a contradiction to make people pay for the right to program.


If you have a “near perfect functional alternative” then what exactly is the complaint?

Demanding all software be free is also demanding the end to freedom. People want iOS to be more open because of the incredible value of iOS. Not because there’s an equivalent free alternative at hand they simply didn’t notice.

> Darwin/XNU is libre software, so it's a contradiction to make people pay for the right to program.

This does not make logical sense? That other free software exists is not an argument that all software must be free. Apple has a business model which increasingly relies on selling services and licenses on iOS over selling new hardware. That’s their choice for how to fund their operations, which I’m very happy that they are free to make!

Being forced to make all my code freely available would be an appalling restriction on my own personal freedom. Not to diminish the brutal history of slavery, but what gives someone the right to free access to my work?


>is the complaint?

That Apple's investment of billions into the ecosystem is not an argument I see valid, given the alternatives.

>all software must be free.

Heavens, that's not what I'm saying. I find it contradictory that Apple have open sourced an entire operating system and an entire kernel, and are kvetching over a mere privilege to begin approaching a distro of their OS. I'm not fighting the freedom fight, I'm in awe of the bait-and-switch they're employing.


I think they are trying to protect a billion dollar ecosystem which is their App Store.

Apple's App Store earns developers something like 50%+ more revenue than equivalent apps on Android. No small part of that is due to the ease of piracy on Android.

I do not see any contradiction, nor any "bait-and-switch". The operating system and a kernel are not the services engine which keeps their company running. Open sourcing that code doesn't imperil their primary revenue streams.


The very foundation their ecosystem runs on is free and open source, and up for grabs for forking and modification. However, their proprietary app store is not, and neither is force-loading my own custom apps. In my point of view, this is where I see the contradiction.

Granted the worldwide open source community has had and has nearly zero interest in making a darwin distro with a minimalistic tiling window manager, or anything similar. Yet I see gnu/linux developers create dozens of linux distros and dozens of functionally identical window managers. So much for side-rant.


That hasn't been the case since 2016, when Apple changed it so the fee is just needed to publish to the App Store.


No, it's also needed to install a self-made app on my phone for longer than 7 days at a time. That 7 day restriction is far too low to be reasonable for normal use.


"You shouldn't be concerned about freedom because you're not gonna use it/need it anyways."

That's what your comment boils down to. You really want to go down that road of an argument? On a website that has hacker culture in its very name?


Hacker News doesn't really have anything to do with the hacker culture, except yes, having the word "hacker" in its name


You can run anything you want, but you can't not run anything you want.


Don't forget about Telcos.

Your AT&T, Vodafone, Verizon & Co, put an enormous amount of pressure on Apple to limit and protect device unlocking.

This is a cat & mouse situation where either party might benefit from tight unlocking controls.

For example, MVNOs benefit from manufacturers with flexible and open unlocking policies that make it easy to unlock devices (without approval from the original operator), whereas long-term commitment contracts with traditional operators want to make it as hard as possible for you to leave with an unlocked device (regardless of whether you are legally entitled to or not - but that's a different story).

The economics of whether Apple would benefit from less strict unlocking policies (ignoring Telco's wishes) are not that clear.

You might think that unlocked devices would have longer lives, and therefore limit Apple's ability to push a new one to you, but you could also say that giving phones 2 or 3 different owners in their lifespans could help with app store purchases, limiting jailbreaking and possibly avoiding going with a newer less expensive android version.

Disclaimer: We unlock phones for MVNOs and individuals who get bullied around by their telcos.

Edit: Spelling


Apple sells plenty of unlocked phones, all you have to do is buy it from Apple.

Lock restrictions only come into play if you want the mobile network to subsidize your phone.


Agreed! But a big majority of phones out there, especially in secondary markets to the US like Latin America are locked by default from within the Telcos, and will go though several hands before ending on a shelf or refurbished and sent to the middle east.

Each country has different policies when it comes to unlocking. Locked phones greatly outnumber unlocked ones.

Take Chile, where phones must be legally unlocked by the telco or manufacturer to work WITHIN other operators in the country, but not necessarily abroad.

Buy an "unlocked" phone from a Chilean and you might get stuck with a brick.

The different and complex type of unlocking levels Apple has for its devices (and there are many) are designed to assist the Telco and any countrywide regulatory policies that need to be enforced.

There are many parts to this jigsaw puzzle.


Yeah it’s more law related than anything.

Here in the France the telco are compelled by law to fully unlock any phone bound to their network 6 months after the purchase (or earlier if initial contract duration was lower than 6 month).

Apple comply gracefully to this law.


Or, you know, buy something else.

Before the"locked devices, people had the inverse problem:

everything was two open ended and complicated, could cripple the system, stuff was open for exploit (much more so than in this case of unpatched vulnerability, viruses were everyday occurence). Techies didn't have this issue, but the general public did (heck, even techies did suffer somewhat). And that might have been OK for when the PC was an unconnected tool, but not when everybody has one with them, from school to nursing home, and everybody buys and does tons of sensitive stuff from it, while connected 24/7.

Whereas you can give a 2-year old an iPad, and they can start using it just fine...


Give a child an iPad, and they’ll be bombarded with toxic advertising and apps wanting money, money, money.

Back in ‘the old days’ (80s+90s), a child with access to a computer was likely to learn something, even if their primary use of it was playing games.


>Give a child an iPad, and they’ll be bombarded with toxic advertising and apps wanting money, money, money.

Every been to the modern web? Or used 90s-00s shareware?

Compared to those, the iPad is advertising and nagging free...

And can be totally free, you don't have to buy (a) adware apps, or (b) games with in-app-purchases.

I only get stuff that's in neither category, which all the best apps are...


> Whereas you can give a 2-year old an iPad, and they can start using it just fine...

What makes you think that if Apple would provide a mechanism in iOS, like SIP (System Integrity Protection) on macOS which can be disabled by technical users if needed, would change anything?

Right now, regular users do not disable SIP on macOS so there's no issue there. They don't even know/care that this possibility exists. So your 2 year old could still use the same iPad just fine.


> Whereas you can give a 2-year old an iPad, and they can start using it just fine...

That's because this new devices aren't "secure" but severely limited and crippled, you can't do much with them and they are far from actually usable like a computer. By that metric, my old Nokia was even more secured than an iPhone.


>That's because this new devices aren't "secure" but severely limited and crippled, you can't do much with them and they are far from actually usable like a computer.

That's the whole point: for them to not be as open ended and complex as computers, while having the power to run highly feature-full apps.

>By that metric, my old Nokia was even more secured than an iPhone.

It indeed was, and that's the ideal. To make extremely feature full modern smartphones as easy to use and as complexity/trouble-free as appliances...


> everything was two open ended and complicated, could cripple the system, stuff was open for exploit

Maybe on Windows, but on macOS this has never been generally a problem.


> general computing

Apple does not consider phones devices for general computing and so prioritises stability, power consumption and security over flexibility and the ability to run arbitrary code. I'm happy with that trade-off.


They clearly consider iPads general purpose computers, so we're back to the starting concern.


I have never thought that Apple considered iPads general purpose computers. One of their biggest ads about the iPad Pro[0] even implicitly states that.

The proposition, in my mind, that Apple is trying to sell with the iPad to the majority of people who use their computers for social media, content consumption, and office tasks like email and Word document authoring is "you don't need general computing flexibility for the vast majority of things that you do".

[0] https://www.youtube.com/watch?v=llZys3xg6sU


This is such an unfathomably incoherent thought from an individuals perspective that i don't even know where to start.

Could you give your definition of 'general computing'?


Well, I can run a Turing Machine using javascript on my phone's Safari browser. Do you have another definition of 'general computing' ?


Which part of it do you find unfathomably incoherent? I'll try to explain very simply.


Steve Jobs said in his keynote of the very first iPhone, Apple have put a full-featured operating system on it, rather than a baby OS.


I'm not sure Google really belongs here. Telcos and some companies licensing Android - sure. But unlocking actual Google phones, like Pixel, is literally available from the menu in the developer settings. It's a well known/documented process.


That only unlocks the bootloader, which then allows you to install a custom one. There is no built-in option on Android for gaining elevated priviledges in any form, not even in a restricted way to disable bloatware that might compromise your privacy, and rooting your device breaks OTA updates, which also include security patches.


It does. Unlocking and gaining elevated privileges through rooting trips Google's SafetyNet API which will lock you out of many apps and/or features. It also breaks your device's Widevine certification so you can't watch DRM-protected video from many services.

As for the "We do not control our own devices, we cannot stop certain processes on them, and we do not know where our personal data is sent.", with stock Android provided by an OEM, you have no control over the opaque and invasive monolith that is Google Play Services and there's no way to control what data exits your device, aside from installing a VPN-based firewall. You can't even control when apps have access to your data, just a binary on/off switch.


Of course it breaks DRM. That's the point.

> with stock Android provided by an OEM, you have no control over the opaque and invasive monolith that is Google Play Services and there's no way to control what data exits your device, aside from installing a VPN-based firewall.

This is true of all systems (and even worse on iOS, which has the same data collection as Play Services but can't be disabled or avoided) and irrelevant to the discussion about whether a user can completely control their device if they wish.


That's a different question. You own the device. Now you are demanding that other services interact with your device regardless of how it is configured.


It's fortunate I can buy an iphone for my mum and she hasn't had a single problem with it.

On the other hand I bought my dad a laptop with Windows and despite having an antivirus he's had all kinds of problems with it, including some heavy duty adware.


Indeed, I wouldn't suggest anything other than a chromebook for my relatives


You can setup Windows to be secure, just make sure to not give your parents admin accounts. Install the apps they use and give them normal user accounts.

Provide the password for elevation and educate to only type that in when they are installing something they know is secure. Likely they will forget the password and have to check with their "IT helper" anyway, and sanity check there actions then.

I have a Windows 7 machine setup like this for my parents and they have never had a problem with malware. The stuff that lives in the user profile gets caught by AV, and I have to install something for them maybe twice a year.

I would rather put in a little extra work setting up a Windows laptop than send metrics for the entire system to Google. Their Android phones take care of that invasion of privacy.


You don't need admin access for anything. Without admin privileges I can install new software, sniff your passwords, encrypt your files, participate in a DDoS attack, mine Bitcoin...


> You can setup Windows to be secure

Or I could get them to buy something far cheaper that's already secure, rather than having to learn how to admin an OS that I have had no need to use for nearly 20 years.

> check with their "IT helper"

The whole point is to avoid being an "IT Helper", otherwise I'd just give them an ubuntu laptop.

> The stuff that lives in the user profile gets caught by AV,

Audio video?


Do you consider https://download.lineageos.org/ a questionable source?


I do not, but you have to be lucky enough to have a device that is supported by LineageOS, or you need to spend time learning the skills to build a custom ROM for your device.

Why isn't there an option in developer mode that gives us a root shell on our Android devices? Why is an escape hatch that gives back control to the user so frightening for these companies?


You don't need a custom ROM to get root on Android, you just have to unlock the bootloader and replace the "su" executable. Moto (aka Lenovo) devices can be unlocked without exploits using their online tool, for example.


I mean, it's not luck if you just buy one of those devices...


It's because that same escape hatch can be used by bad actors.

And unfortunately as we have seen all too often users are willing to blindly do what they are told if they get something out of it e.g. free game, credits etc.


I don't think that's so bad, makes the developer's job easier when there is just one version of your target. Nothing worse than entitled tinkerers.


For a second I thought I was in a thread about Windows


There are things I want to control, and there are things I do not want to control. Smartphone is not a "general purpose computer" and I want just use it not to babysit it. And more often than not having "control over your own" device means that is is just some malware that has this control, not the user.


Your smartphone has all the hardware that it needs to be a general purpose computer. If you could plug in a decent screen and connect a mouse and a keyboard it would be more powerful than the computers you used just a few years ago.


So does around twenty things in my household: My TV, my camera, my router, my wifi access point and perhaps a dozen more. Probably 3 or 4 I didn't even realize. That doesn't change the fact that I don't want to know and care about that. They are appliances, meaning I prefer a locked down system if it means they are simpler and more secure.


>Your smartphone has all the hardware that it needs to be a general purpose computer.

Which is neither here nor there. Parent doesn't mean a smartphone doesn't have the cpu power etc. of being a general purpose computing, he says it is not one.


But it has the capacities to be one if you plugged in peripherals.


Again: that doesn't matter. My lawnmower has more power than my first 3 PC's combined and could easily play quake if connected to a keyboard and screen. Not sure what difference that makes to the original question of openness?


The point is if it were open it could do that and then it'd be a general purpose computer.

People were justifying locked down smartphones by saying they're not GP computers.

Factually, if they were open, they could be GP computers. The only reason they aren't is because they are locked in the first place! That justification for locking phones is based on a consequence of them being locked, so it doesn't make sense.

That's the difference it makes.


I disagree. I want my phone to be an appliance, just like my router and TV, even though under the hood all 3 have almost the same hardware (all 3 are ARM Linux machines). I want them to be as simple as possible, and achieve that by minimizing the control surface I have on it. My computer (that has a huge control surface) has that because I'm ready to pay the added overhead to complexity/security - but I don't want that on my phone.

"It's not a computer" by that I mean "I don't want it to be a computer" in the sense that if it had the complexity/risk of a PC I wouldn't buy it in the first place.


That's just the side you take in this openness/simplicity tradeoff and not a good reason for saying smartphones should be locked down in general. There are going to be people who want more open phones or more locked down desktops. Users should have the choice not only with their computers, but also with their phones.

That Apple and Google have taken that from the user is unfortunate and I think there's not much of a case to be made against that. All you've been saying so far boils down to: you don't want an open phone, and because Apple and Google have coincidentally made the same choice for their customers this kind of paternalism is fine by you. Kind of a short sighted position, no?


No I mean it’s the obvious position to take when 99% of customers want it. For all manufacturers. If it was just 80% of customers who wanted it then it would be unfortunate if no manufacturer positioned themselves to cater to the 20% who want a “pc phone”, but since those who want it are (surely?) not even a rounding error in the statistics, I can’t see why it’s unfortunate.


I've never understood why iOS isn't more like macOS. There aren't really any technical reasons why this couldn't be.

It's common for the security argument to be used to justify Apple's practices, but Mac users have been perfectly fine installing third party apps such as Transmit, Adobe Photoshop, or even Google Chrome from outside of Apple's walled garden. I've been using macOS since 2007 and I've never had a virus or any security problem, nor anyone else I know using a Mac. From my dev colleagues to my 70 year old mother in law or my 15 year old nephew.

It seems to me the only real arguments for Apple's walled garden are economical.


There will likely be a day when it's very difficult to install anything you want on macOS. Same with Windows. They keep trying really hard to make it happen but, for now, people keep pushing back even harder. For me, the fact that they're even trying is enough to leave.


Market share is an appreciable concern here. Macs haven't been more than 20% of the market for a long time. iPhones are a much larger market share, especially among affluent users. I don't think we'd see any widespread viruses, but there would certainly be a ton of people losing their financial info from their own irresponsibility.


So your argument is that because there are more users it needs to be more secure?


My argument is primarily that if somebody is looking to infect users, are they going to target the 86%, or the 13%? Macs have benefited for years by largely not being targeted by the majority of malware developers. If mobile devices open the walled garden, you get something similar to android's malware issues at best.


iOS does not have a 86% market share. Not even close. Globally it's closer to 15%. In the US it's closer to 55%.

If you argue that the 86% is the case on certain demographics, then the same can be argued about macOS.


The demographic argument is what I'm going with. MacOS is 18%, compared to Windows at 74% [1]. When comparing the mobile statistics, they're bouncing around 50%. The iOS pool would be a much larger pool to attack than MacOS, and for little financial benefit for Apple.

My prediction: If Apple opened the garden, even for a 'developer only' mode, I would imagine unregulated app stores would go up overnight, with wikihow articles on how to enable and install them, followed by a large amount of the technically illiterate (speaking from experience) trying to get free games, or 'add new emojis'. Users can't be trusted, and if consumer desktop operating systems were designed today, they wouldn't have the freedom they currently have.

[1]: https://gs.statcounter.com/os-market-share/desktop/united-st... (Not entirely sure on the accuracy/gathering of this data, but it had the easy filtering and seems to line up with both of our data points)


> followed by a large amount of the technically illiterate (speaking from experience) trying to get free games, or 'add new emojis'

I agree some people would do it... but I doubt it would be such a large number. Do you have any data about this?

Users can already do that in Windows, macOS, and Android. From my anecdotal experience very few do it.


> I agree some people would do it... but I doubt it would be such a large number. Do you have any data about this?

No, nothing more than anecdotal. But I'd argue 'some people' when you're operating on the scale of the hundreds of millions-billions that Apple operates at, would still be hundreds of thousands of support requests from users who have inadvertently made their phone unusable (using some jailbreak tweaks as a reference) to extremely annoying (referencing many android apps that abuse push notifications for advertising).

> Users can already do that in Windows, macOS, and Android. From my anecdotal experience very few do it.

I used to work IT for a school system, and I had an entirely different experience. Teachers would occasionally ask for help with their personal laptops (without our AD) and they were near universally a minefield of toolbars and adware. There's obviously some self-selection in there, but if the 5% of teachers couldn't handle a computer responsibly, that would be a big problem for Apple if they added more ways for users to screw themselves.

Of course, all this is anecdotal, so I'm not expecting this conversation to really convert either of us.


I work in edtech and I can only agree that teachers are some of the more tech illiterate users I've ever seen. :)

But yeah, without any solid data we will both stick to our anecdotal experiences.


I think the argument is that if an OS has a tiny market share then it tends to be more open to 3rd party developers as a way to increase market share.

MacOS used to be a more developer friendly.


The saddest thing about this is the article presents jailbreaking as an entirely negative thing, while the comments offer a more balanced opinion.


For most end-users it is. It allows any compromise of the device to elevate to root privileges.


I'm just thankful for a working jailbreak on the latest version with no "Please update your iOS beta version" popup. It's like a dream come true.


Absolutely incredible to have a jailbreak on the latest signed version of iOS.


No regression tests? ಠ_ಠ


It looks like a bad merge in which if there was a test it would've not been merged either.


(again). Goto fail comes to mind.


I mean... does every single patch need a regression test? If I did free(p); p[i] = 1; and then I fixed it by doing p[i] = 1; free(p); do I really need a regression test to trigger the dumb use-after-free I'd introduced?


I think the answer to your question is not obvious. Here, it would have prevented the problem of Apple. 12 years ago while working on a military project on sun, I have encountered a similar vulnerability caused by a regression https://blog.erratasec.com/2007/02/trivial-remote-solaris-0d...

Adding this kind of non regression test is costly, but it protects against source code management mistakes.


Unless, of course, the code management mistake that removes the patch also removes the regression test!


> I think the answer to your question is not obvious.

You mean the answer could be "yes, every single patch must have a regression test"?


>A third security researcher, Stefan Esser said that people should be careful what apps they download from the App Store right now. "Any such app could have a copy of the jailbreak in it," he wrote on Twitter.

Seems a bit overblown when there's a review process in place. I'm sure it's not infallible, but still..


> Seems a bit overblown when there's a review process in place. I'm sure it's not infallible, but still..

Well, until now I assumed Apple would test for patched vulnerabilities in new iOS releases too.


It's happened before, and it's probably not super easy for Apple to do static binary analysis to determine if an app is going to make exactly the bad syscalls necessary for the jailbreak gated behind an undisclosed trigger functionality.


Can a review really check for this? Aren't there obfuscation methods and ways to delay the trigger of the exploit? It could wait for a special network package to start the exploit.

For iOS 9.3.3 there even was an app in the App Store, that could jailbreak your phone. (PG Client)


Yes, to some extent people should be worried about apps potentially containing exploits, but then again they should be more worried about 0-days than a known vulnerability.


> A zero-day (also known as 0-day) vulnerability is a computer-software vulnerability that is unknown to, or unaddressed by, those who should be interested in mitigating the vulnerability (including the vendor of the target software)[0]

Why should a publicly known unpatched vulnerability be a lesser concern than something that you don't know exists?

[0]https://en.wikipedia.org/wiki/Zero-day_(computing)


Please excuse the tinfoil hat - is there any chance that this vulnerability was reintroduced at the request of the Chinese government to allow easier access to Hong Kong protesters devices?


Cool! My first negative post. Maybe more context is valuable.

We know HKers are getting more and more tracking from the mainland government and are switching to applications not typically used for communication and organization. https://www.businessinsider.com/hong-kong-youth-using-tinder...

We also know that Apple has accommodated Chinese government requests to increase their ability to monitor communications: https://boingboing.net/2018/05/21/apple-bends-to-chinese-gov...

And as one of the most valuable companies in the world that ships software, it is surprising to see regressions in their software.

The article said that right now legitimate applications on the App Store could contain jailbreak code.

Of course these things are entirely possible to be unrelated, which is why I mentioned the tongue in cheek “tinfoil hat”.

The timeliness of the response from Apple in patching this reintroduced vulnerability should be swift. If so, I’m happy to step away from this position entirely.


No need for a tinfoil hat or this specific vulnerability.

Any number of backdoors can be introduced with any update for any operating system or app. Generally governments around the world want backdoors and information from companies and companies generally comply.


Last time this happened to Apple they fought it tooth-and-nail and at least from what the public knows they were victorious and did not have to add a backdoor.


no


Apple has definitely accommodated the Chinese government’s desires before: https://9to5mac.com/2018/05/19/apple-cracking-down-on-callki...


During the San Bernardino case, they have said multiple times that they cannot install software without the user entering their password.

And even if you assume a publicly traded company lying outright, what do you think is more likely: then silently backdooring the phones of people in Hongkong or pretending to have accidentally unpatched a vulnerability, hoping people would update to that OS in order for pretty obvious exploits being able to run?

If I was as evil as you make Apple out to be I would sure as hell do the former.


I got a lot of flak here recently for suggesting that maybe security researchers shouldn't be publishing PoCs or deep vulnerability details literally 1 week after the vendor issues a patch.

Here's to hoping that, now that this happened, someone will give this idea another consideration...

(P.S. for those wondering: apparently this is CVE-2019-8605: https://bugs.chromium.org/p/project-zero/issues/detail?id=18...)


But it's been 3 months since the vendor first issued a patch!


I mean, I'm not suggesting 1 week should've been 1 month or even 3 months. Those are too short to me too.

But regardless, that's already 3 months people had to design, write, test, and perfect an exploit for it...


The exploit was patched in iOS 12.3, not known since 12.3. Apple probably knew for longer, fixed it in 12.3 and reverted the patch (somehow) in 12.4.

If you want to make your point, this is one of the worst examples you can take as it is an old exploit, which has been patched and now works again. The code should be in the public after the patch anyway if a researcher found it.


> The exploit was patched in iOS 12.3, not known since 12.3. Apple probably knew for longer, fixed it in 12.3 and reverted the patch (somehow) in 12.4.

Huh? Am I misreading the timeline? iOS 12.3 was released May 13, and I see the view restriction removed (Label:-Restrict-View-Commit) on May 20... which is almost exactly 3 months ago: https://bugs.chromium.org/p/project-zero/issues/detail?id=18... https://support.apple.com/en-us/HT210118


Gives people opportunity to jailbreak their devices?




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

Search: