Apple, I get it. I get why you do it. But seriously, this retroactive punishing of prior 3rd-party purchases is asinine. I had purchased multiple lightning cables from Anker (a very reputable brand), and had one of the iOS 11.x updates basically render them useless. Apple's product support is not fast enough, nor pervasive enough (to say nothing of cheap enough), to justify this behavior.
It was not done intentionally. They simply don't QA their releases on repaired phones (because why would they?). The touchscreens in question highly likely use an older touchscreen controller from a previous iPhone. These controllers don't support certain commands that evidently they are now using.
As far as the lightning connectors go, that WAS done intentionally. The third party lightning connectors generally don't implement a very specific charging protocol designed to prevent metal migration between the pins and the pads on the connector. They only implement the handshake. But guess who has to honor the warranty if you've used a non-compliant charger cable?
It'd be pretty hard for them to know though. They'd have to reverse engineer (the irony isn't lost on me) every variation of third party screen/touchscreen controller. What happens if some feature isn't available on a third party screen? Either they don't implement the feature at all or third party screens are now worse off and we'll have a headline "Apple cripples third party repairs!".
This is precisely the kind of stuff that Microsoft is famous for, and it is entirely possible. Heck, third party hardware is what made PCs great. I think, however, that there is a difference in customer base; Microsoft caters to businesses, Apple to consumers.
When Microsoft breaks a third party app/device with a Windows Service Pack, Microsoft needs to fix this otherwise their customers get angry. When Apple breaks a third party app, they require the customer or third party app provider to fix it, because Apple is inherently a company that wants to control the whole stack and doesn’t embrace an open ecosystem.
It's amazing that I find myself thinking that Microsoft of the 90's and early 2000's as actually being a more open and accommodating company and windows being a more open and accommodating environment than Apple and iOS today.
Whether Microsoft had no choice and would have been worse if they could in this respect, the fact they they weren't and I think Apple is, and Microsoft was widely lambasted as money driven (e.g. M$) and evil at the time, and Apple largely gets a pass now, is interesting. That could be because Apple is better at managing image, or because people's perceptions and what they expect has shifted, or because there's a lot more bad (and worse) actors out there now that make Apple look better in comparison. It's hard to tell exactly why it's looked upon more favorably now, but I do believe my thoughts on Apple to be accurate, and I'm not sure how to feel about that.
Fruit company pulls the plug on anyone who writes anything bad about them. No more marketing money, no more freebies, so publishers think twice about it.
There's a major difference here: a 3rd party computer part drawing 5% more current than expected probably won't cause issues (There are however many of cases where they do...) whereas a component drawing 5% more current on a phone may cause a shutdown.
When Microsoft broke third-party stuff, the end result didn't cause fires. It was also done very deliberately to damage competitors, according to information revealed at trial.
Software companies can do things a bit differently from hardware companies.
Can you imagine the maintenance nightmare in having special case hacks just to support bugs by one program? I read some of the MS blogs where they detailed all of the workarounds they've had to do. It has to make Windows less stable and harder to maintain.
OR provide third parties with API info, maybe even the third party implementation. It doesn’t even have to be official. Apple knows very well how to leak controlled information. But it would greatly improve the quality of third party repairs, and lower the load on official support workflows.
Slippery slope here. If they start testing with 3rd party hardware, that market will grow for third party parts whereas if they constantly tell people they don't support it and casually break those devices by accident, then they curb the issue while it's a small segment of the market.
I see a company like Tesla taking the same approach.
If we keep going down this slippery slope, next thing you know there'll be stores called Phonezone or Advanced Phone Parts on every street corner selling repair parts for every popular brand of phone at reasonable prices enabling consumers to do their own repairs at home.
I’ve worked with many LCD panels: while they all support one of a few signaling standards, there is no standard for the cables and connectors. (Or if there is, nobody follows them.)
For each LCD panel, you need to make a custom cable.
So that part is not something that’s specific to Apple.
"A similar thing happened for the iPhone 7 last year. An iOS update prevented the touchscreens from working on iPhone 7s with third-party repaired screens. Apple then released a follow-up software update that made them work again, resolving the issue."
To me, that makes axoltl's explanation far likelier. From Apple's perspective, I get it: they do Q&A on what's officially supported. Trying to pull in all of the unofficial changes is more expensive and potentially restrictive.
From my understanding, many Apple chips are custom, and I assume much of their firmware is too. What your asking is for Apple to support third party knock off chips in their firmware and drivers which were most likely reverse engineered to "get it close to functional". That's a bit silly.
Imagine getting upset that an Occulus firmware update didn't work with a knock off HMD that lacked features of the original...
> What your asking is for Apple to support third party knock off chips in their firmware
They're already doing that for some other things.
For instance, some (mostly cheap) Wifi access points are known to have buggy implementations of 802.11 that require firmware workarounds in phones and laptops. Manufacturers have to support these devices for the benefit of the consumer and to maintain their own reputation. Customers aren't expected to understand that it's some other piece of tech that's at fault.
Granted, it's a bit different since this is a separate device. But this is still one example of Apple (and every other phone manufacturer) supporting other's people buggy code for the benefit of the users.
My previous life was network hardware validation, so I have some perspective on this.
What your describing is "interoperability testing", and it's ubiquitous in the communication standards world.
Standards sometimes aren't clear, sometimes there are multiple implementations of a feature before a standard is ratified (and a big user base with the now non-standard implementation), and sometimes bugs end up in silicon and firmware.
One company I worked with had a very large lab with hundreds of ethernet cards. A column of robots would cycle through each one and plug a network cable in, make sure the link was up and solid, then move on to the next.
I think this is fundamentally different than Apple supporting third party hardware.
Communication standards are designed to allow chipset a to talk to chipset b. Both have an obligation to get a link up. A failure to get a link up is a failure for both, from the customer perspective.
A closed, proprietary, custom hardware system has no such obligation. Unlike the communication channel, the risk is completely unbalanced. A failure for the knock off hardware design is just directed to Apple (as you see in the comment section here).
Sure, you can't really blame the customer since the knock off parts aren't easily detected or may have been installed by the previous owner. You also can't blame Apple since you've made a custom iPhone with mystery parts, and I think it's unfair to make them stop updating firmware/software so they don't break these mystery parts.
I don't believe the onus is on Apple to /test/ for all possible permutations since they own and control every component of their hardware, but the onus should be on them to fix any phones rendered inoperable by another iOS update unless its to prevent a bigger issue, 3rd party batteries exploding, etc.
This is also another good reminder to postpone the iOS updates as long as possible; I think it's been since v10 when the updates have had a less than stellar record.
> but the onus should be on them to fix any phones rendered inoperable by another iOS update
How is it their responsibility to support third party hardware? And, where's the line for support?
If I make a very poor knock off screen with a design issue that is made apparent with an iOS update (say, incorrect implementation of a required spec, interface, timing, etc), why should Apple have to support a flawed design?
What if Apple included a feature in the hardware that but the software wasn't ready at release, but would be made available as a standard feature in an upcoming iOS update? Should Apple have to continuously fragment their codebase, resulting in some tree of hacks to support each new half-baked chipset?
To someone in the hardware and software world, this perspective is absurd.
I find this stance very odd. How do you imagine them achieving this? Do they have someone constantly ordering screen from the hundreds of vendors on Alibaba, in some attempt to make sure they "find them all"?
If you could give one example of a company with a custom and proprietary system and chipset supporting third party knock offs I would be very interested. I've never witnessed this during my time in the hardware industry.
They probably can (and I guess rather easily) test whether the touchscreen is or not from Apple and at least let the ungood ones work as before (I am sure this is pretty easy to do: just keep the old controller), issuing a warning (not that Apple is short on these) to the user: "your screen may malfunction or may be insecure. If you had it repaired outside the Apple network, we remind you that it is not covered by our guarantee".
It's neither easy nor cheap. The engineer in question now has to support multiple variants of the driver, and QA has to acquire several of these screens and maintain a dedicated test rack of them.
I'm a little amazed by how many people in this thread don't understand how much work developing for, testing, and maintaining actual physical devices is.
> The engineer in question now has to support multiple variants of the driver, and QA has to acquire several of these screens and maintain a dedicated test rack of them.
Conversely, the update just needs to reliably detect if the screen is Apple brand, and abort the update if it's not. If this can be done, then the current driver could be considered generic and working everywhere, and the engineer would just need to maintain the generic and the current version.
That is, if Apple didn't mind to pay for making easier for third parties to get sales instead of Apple. Somehow I don't think that's the case.
The update is an all-or-nothing type of deal though. You either update the entire system or you don't. So now people with a third-party screen don't get new features and - more importantly - they don't get security fixes.
If the update wasn't all-or-nothing, now you have to test interoperability between this generic driver and the rest of the system. That might be OK if it was just this one driver, but what if you suddenly have a generic driver for N subsystems? The combinatorics explode quite quickly.
Even if it's all-or-nothing and no generic drivers are implemented, it's better to deny new features than to cripple a working device. Apple defines the third-party market with how much documentation it releases. This is not a technical or practical problem, it's Apple not easing its iron grip, as it always does.
I looked at it, but that doesn't actually tell me the touchscreen interface. It just tells me the bits they're currently using. If Apple had open sourced their driver it wouldn't have made a difference, as whatever they're doing differently wouldn't have been in it until now. In order for me to make a fully compatible touchscreen, I need the datasheet and programming manual for the touchscreen controller.
As much as open source is a thing for software, it is not for hardware and especially silicon.
> I looked at it, but that doesn't actually tell me the touchscreen interface. It just tells me the bits they're currently using.
I have to agree with you on this. This is not a full hardware specification, it just tells how the hardware should behave up until the driver was released.
Modern open source hardware is still far in the future.
Open source hardware's still far in the future, and I'm actively worried about open silicon. The RISC-V foundation is making steps in the right direction, but it'll be a long time coming before I have a document that isn't under several layers of NDA telling me how to configure the memory controller on the latest Snapdragon.
Well, that is an opinion. However, bricking a device without previous warning seems quite out of control. There are many easier things to do. Like, for example, nagging the user every day with a message, not upgrading the software (and still nagging the user with a message), etc...
Bricking a device is certainly a bad user experience and not at all what the user wishes. Nagging him at least lets him use the phone and keeps him informed.
You cannot assume the user is stupid and not assume the user will not repair the phone outside your network. You cannot have both things at the same time and get away with it.
That is what I think. If the user is stupid, let him be so.
For goodness sake no! Blocking software updates is a terrible thing to do, because most people don't want to install them anyways. You're basically putting people in an already bad situation in a worse situation.
> Nagging him at least lets him use the phone and keeps him informed.
It only takes the user using the sketchy Touch ID sensor once for their fingerprint to be sent out to malicious actors.
> If the user is stupid, let him be so.
That's what you think, but that's not how Apple works. It's their job to make the default work well for those who may not be completely knowledgable about security.
I get what you are trying to convey. I disagree but there is no problem in that.
However: where does Apple state that their products may stop working if repaired with different components? (Honest question, I have tried to find it).
This was that best I could find from a few minutes of searching.
> Service Exclusions and Diagnostic Fee. Apple may charge you a diagnostic fee (including shipping charges) as described in the Country Variation table, below (“Diagnostic Fee”), if Apple inspects your product and determines that (i) your product does not require service, (ii) your product has failed due to or has incompatibilities with software or data residing or recorded on your product (iii) service is required due to the failure of parts that are neither supplied by Apple nor Apple-branded, (iii) additional labor or parts are required that were not specified in the original estimated charges and you do not agree to authorize service based on Apple’s revised estimated charges, or (iv) service cannot be performed because the serial number has been altered, defaced or removed or the product has failed due to accident, abuse, liquid spill or submersion, neglect, misuse (including faulty installation, repair, or maintenance by anyone other than Apple or an Apple Authorized Service Provider), unauthorized modification, extreme environment (including extreme temperature or humidity), extreme physical or electrical stress or interference, fluctuation or surges of electrical power, lightning, static electricity, fire, acts of God or other external causes (“Service Exclusions”). Apple will return your product to you without servicing it and may charge you the Diagnostic Fee.
This is from Apple's Repair Terms and Conditions: https://www.apple.com/legal/sales-support/terms/repair/gener.... It doesn't say that the product will stop working, but it does say that Apple isn't responsible for fixing it if it goes wrong. I'm not sure if that could be extended to them not having to support your third party component in software, though.
Did Joe's Fly By Night iPhone Non-Genuine Replacement Parts disclose that their replacement part wasn't the same as Apple's? Seems like that's where the onus lies.
Apple has a program for third-party stores that use genuine parts, the Apple Authorized Service Provider program. If the repair shop you go to isn't an AASP, then you can't say they have genuine parts.
Except this particular issue has absolutely nothing to do with TouchID or security at all. It's just the screen. Error 53 was significantly more justifiable.
Apple relies very heavily on customized silicon to provide an edge (Lower power consumption, for example, or extra features). This means that there might be a part out there that's sorta close, but not quite. Once Apple starts pushing their customized silicon, the off-the-shelf part doesn't keep up.
It's not like repair shops are doing this to cut corners and costs or cheap out. It's simply impossible to source the exact replacement chips for recent iPhones, Apple doesn't sell them, in fact they probably ask their chip manufacturers for exclusivity - and at the scale of Apple orders, manufacturers are happy to do whatever.
If this is the case, wouldn't the "touch current" from capacitive leakage on the Y class capacitor tying the SMPS ground to mains neutral do this anyway?
Surely, if electromigration is an issue, wouldn't these issues be just as significant?
Isn't gold practically immune to electromigration at these energy levels?
The problem here is that Apple leaves third parties in the dark. A reasonable, open, spec and stance on more complex repairs would shift the responsibility to the repair company.
It seems more likely that Apple is checking serial numbers or some kind of embedded meta-data, to verify that the parts in the phone are authentic Apple parts.
> To mitigate these issues, external contacts may typically include a corrosion-resistant coating to the exposed exterior surface… Although temporarily effective, such coatings are subject to frictional wear and may become less effective over the life of a device.
Frictional wear is not a huge problem in practice.
There’re billions of devices with built-in batteries in the world. I’ve been carrying a cell phone, and later sometimes other electronics, since 1999. Never encounter such issues.
Most manufacturers use gold plating, and they advertise reliability for 10000 mating cycles:
This is the point I'd have made if I had more patience. It is simply ridiculous to claim that galvanic corrosion and/or electromigration is likely to be a problem on a square millimeter gold contact with a few microns of metal that passes at most 15W continuously for a few years before being discarded.
I'm confused how your Anker cables stopped working. Mine continue to work and Anker is part of Apple's "Made for iPhone" accessory certification program.
Anker produces officially licensed cables that legitimately carry the ‘made for iPhone’ mark. If they stop working it is much more likely something else caused it instead of a software update.
What a charger? Does the phone simply not let the power through if it doesn't receive the correct bits from the signal wires? Or was the charger breaking some voltage/amp limit?
The lightning port is a double edged sword. On the one hand, the wires are "smart" and can adapt to any protocol, so the same port can theoretically function as a USB or ethernet at the wire level (again, theoretically). But to make that happen, every plug has a chip that negotiates the protocol when plugged in.
It's likely that negotiation failed in some cases.
That's absolutely crazy for most of us who think "this is a wire, why the hell would it fail?" but this kind of crazy complexity in a 5mm plug is possible nowadays.
I find this even more despicable because Apple chooses to use a certain type of rubber under the guise of it being more eco-friendly.
Yet this rubber is ridiculously brittle and frays almost without fail within 6-12 months meaning you just end up having to by more cables thus using more of the worlds resources anyway than if they had shipped you one cable that didn't fail.
My 17 year old Ti Powerbook cable used daily for 11 years, no fray.
My original iPod cable, no fray.
My iPhone 3G cable, no fray.
My iPhone 4S cable, 5 years use daily, no fray.
Change happened somewhere here, newer cables are more matte and softer feeling
My iPhone 7 cable, frayed within 6 months
My 2014 MBP cable, frayed within 12 months
Replacement cable, frayed within 12 months
So you're saying that you're sure that the problem that people usually attribute to bad strain relief is actually caused by slightly different plastic?
Hard to know if this was done on purpose, or if the third party replacement parts are actually substandard/not-to-spec and therefore naturally not QA'd at Apple.... at least until a technical root cause is found?
If it's anything like what happened last year to the iPhone 7s it's a security issue relating to detecting a non-authentic TouchID button. It may really be a mistake, as Apple may not have a third-party screen at all to test updates with.
Replacement iPhone screens are of dubious quality. I had my 6S screen replaced, the first one I couldn't see with my sunglasses on and the second one the touch screen was very erratic, and the one I have now has terrible light bleed at the top.
the issue is apple refuses to provide third party repairs with parts, or even specifications for parts, so its a bit of a roundabout arguement that is tantamount to saying "only apple should repair them" to blame them as they have no alternative.
Completely untrue - if you go to Apple's own support site to schedule a repair, multiple 3rd party options are given, all of whom can acquire first-party parts from Apple.
This isnt the sale of parts to people that wish to buy them, its a very stringent certification program that very few are going to meet the requirements for, and requires you do things like
"Actively promote the Apple brand as part of their business along with AppleCare service and support products."
"Service Providers are required to use Apple Certified Macintosh Technicians when conducting diagnostics, Covered Repairs, modifications, alterations and upgrades on Apple products."
" Individuals or sole traders may not apply."
Imagine these kind of restrictions being placed on car repairshops before they were allowed to buy and use replacement parts from manufacterers, there would be a massive lawsuit.
Plus, for TouchID enabled phones you MUST have a special rig in order to re-pair the sensor, and for FaceID you have to have a special rig to re-calibrate the sensor stack every time you open the device!
Tesla has restrictions that are similar. Many parts are "restricted" at Tesla will not sell them unless you're a Tesla owner, or certified body shop. I'm not even sure they have any certified general repair shops.
The amp version of this page takes my browser 1.39 seconds to load with about 616KB of data over 27 requests. The non-AMP version takes over 10 seconds, and loads 2.2 MB over 300+ requests.
In this case, it's just a light version of the article with the same content. That seems objectively better. If AMP wasn't incidentally in the URL path you wouldn't have noticed it was AMP. If the existence of AMP encourages people to write faster, less cruft-y pages, that's a good thing to me.
> The amp version of this page takes my browser 1.39 seconds to load with about 616KB of data over 27 requests. The non-AMP version takes over 10 seconds, and loads 2.2 MB over 300+ requests.
AMP isn't a requirement to build small, fast, performant websites. The fact that The Verge ignores those qualities with their non-AMP website is on them.
I understand this thread has really gotten off topic from the original, but I think it's a good discussion nevertheless. I think the criticisms we often hear about AMP tend to be more principle-based/philosophical in nature.
Yes, the overall user experience is better. It's not that the experience is bad, it's more about the means they're using to reach those ends.
I see AMP as a set of training wheels that Google was able to get publishers to adopt. All the tools were available for publishers and their developers to make performant/clean sites to begin with, they just chose not to do so. AMP is basically a standardized form of "the way they should have built their site in the first place".
My other issue with AMP is that it makes companies focus on performance for mobile, but not for desktop. Why not build a quality, high-performance site for all platforms?
At the end of the day, is AMP a net positive? Maybe, I'm willing to consider the end result, but I don't think it's a problem for us to explore and criticize the means or the reality that necessitated AMP's existence.
This seems like a potential violation of the Magnuson–Moss Warranty Act, which prohibits the refusal of warranty service for using third-party parts without proving that the third-party parts were the source of the malfunction. Assuming Apple will refuse to repair these under warranty, which they almost certainly will, this is effectively requiring the purchase of screens from Apple to receive warranty service. On the other hand it's somewhat debatable whether the screens or the OS update are responsible for the problem, but the screens were clearly working fine before the update.
Given that iPhones with OEM displays are apparently still working fine...isn't it pretty obvious that the third-party parts are the source of the malfunction?
There is zero evidence so far as to whether this is intentional sabotage of third party parts or not. However Apple has already demonstrated that they are able and willing to do this kind of thing i.e. with non-certified Lightning cables that also previously worked perfectly fine.
If there is a bit of code in the OS update along the lines of
if(DisplayID != GenuineApple) { breakPhone(); }
then it's obviously Apple's fault. But it's not at all clear yet exactly what's causing this problem and whether or not it was intentional on Apple's part.
I replaced my iPhone screen (at a Staples office supply store) and the new screen was polarized such that it was dark when viewed (in portrait orientation) through my polarized sunglasses. Very frustrating that I had to either look over my sunglasses or rotate the phone to use it. Apple's screens appear to have some slight polarization but it is cleverly at a diagonal to both portrait and landscape orientation. Details matter.
Apple added circular polarization display tech way back in iPhone5 or 6 I think. I used to notice this in my iPhone4S when I had my polarized lenses on.
This could have been entirely innocent. It happened to an android device I am familiar with as well. Third-party screens will usually identify over i2c/spi/whatever the same as original, but you cannot apply original firmware updates to them (they are different internally and will get bricked). However, as they are designed to appear original, there isn't always a way to tell. The two options both suck: do not ship firmware updates (bad) or ship them and risk damage non-original screens (marginally less bad). The choice in the case I know of was made to not ship update (it was minor). Apple chose otherwise.
OS performs a handshake with the target subcomponent and authenticates that the component really is manufactured by the manufacturer that is targeted by the pending firmware update. If the authentication completes successfully, then push the firmware.
A flip-side alternative is for the component to only boot firmware that it recognizes as signed by the correct authority.
In the first case the device has an embedded identity within the hardware. In the second case the device merely needs to validate a signature. Crypto acceleration is becoming very widespread and very cheap so I don't see either of these as difficult to manufacture.
You've obviously never been involved in the design/manufacture of an ASIC.
What you're proposing will take roughly 2 years, and be an organizational nightmare. 'Just' having an embedded identity is already complicated:
1. Where do you store the identity? Fuses in the ASIC? Now you need a fuse bank. Not every process node supports fuses, so you may now have to port you entire design to a new process node.
2. Using this new-fangled identity means you can now perform a handshake with the host. Let's 'just' put down an ECDSA accelerator. What do you mean, this increases the die size by 33%? What do you mean it needs to be resistant to differential power analysis? Oh, right, because stealing a single identity means you can make as many clones as you'd like, and we can't revoke identities if we think they're stolen because of laws in China!
Etc. etc. the second option doesn't even work because the cloned device will simply ignore the signature on the firmware.
I used to think this was all trivial too, but having been through this _exact_ wringer 4 or 5 times before I can tell you it is Hard(tm).
I most definitely have been involved in ASIC design. I appreciate how difficult it is. But, the time has come to recognize parts authentication via hardened identity to be a requirement in pretty much all ICs.
Your first suggestion is what is being done. The parts are pretending to be authentic. This game has been played for decades in various kinds of high-volume computing parts without a good solution.
Similar issue with the 6S happened when ios 11 was released. In that case Apple DID allow people to downgrade (back to 10.3.3 in fact!). I don't have a third party screen, but I did take advantage of that and tried ios 11 for a while. Eventually went back to 10.3.3 before that window closed (they fixed the issue with 3rd party screens on 6S).
iPhone 6S on 11.3. I just replaced the rear camera assembly and the front camera+light sensor+earphone assembly with third-party parts (on a moving train, I might brag). Everything works except the auto-brightness setting disappeared. I did Erase Settings and it's
still gone. I'm wondering if:
- I didn't place the light sensor correctly
- I touched it/ruined it
- the part is defective
- Erase All Content & Settings options is absolutely necessary
You changed the premise. What's being made better? The change being discussed is broken phones. That's not better. Hypothetically, that could be from a change that makes something better. Is there any reason to believe it though?
Fun fact, it'd actually be illegal for Apple to disable devices if it detects a replacement part.
There are anti-trust laws against 'tying agreements', and forcing consumers to only buy components from Apple (due to tie-in) would be violating those laws.
That said, those laws don't say anything about having to interoperate with an inferior part.
So assuming Apple isn't willfully violating anti-trust laws, we can be fairly sure the change was intended to improve some aspect of the touch controller.
Note: There are exceptions if the tying serves a purpose other than maintaining a monopoly (such as the security pairing between the TouchID sensor and FaceID camera).
I had my rain sensing windshield replaced. The insurance company refused to replace with an OEM windshield. The windshield company said it would be fine.
It was not fine. Rain sensing didn't work again for the rest of the time I owned the car.
You can do whatever you want to it! But would you expect an NVidia graphics card to work with AMD drivers? The controller isn't the one the drivers were written for, so it breaks. Not surprising.
I think this analogy has gone off course a bit. But I do think a manufacturer providing automatic irreversible updates has an obligation to ensure, as best as possible, that the device continues to work with that update. If it worked yesterday, it should work tomorrow, period.
It would be perfectly reasonable to not force updates and allow users to continue with the current version forever if that keeps them running. But it's not reasonable to just break people's functional devices with an unavoidable update.
How would the manufacturer know you were using a non-OEM part, though? As far as they're concerned their parts support so-and-so protocol, so they should be able to use it.
This isn't a black and white issue. Some degree of compromise is needed as neither side is 100% right. What seems reasonable is a degree of resiliency be built in - so if the protocol worked on a subset of features, you don't brick it as you expand to new ones, you merely degrade.
It's not a perfect analogy, but the rules on radio interference with devices spring to mind: you need to cope with some interference and you need to not emit it. This results in a more stable ecosystem, even though you could sensibly argue that if one side is stuck to absolutely, the need for the other wouldn't exist.
They don't and shouldn't. But they should and can detect if a pending update will break your device, or downgrade more gracefully when their software runs on hardware that is missing some expected features that were previously unused.
I assume this wasn't intentional (Hanlon's razor).
But even if it wasn't, it's not entirely outrageous to predicate continued software support on some terms. The only problem then would be not making those terms clear up-front.
Edited for clarity