U.S. To Probe Apple Over Updates That Slow Older iPhones
bloomberg.com
bloomberg.com
Apple could be inserting code to slow down phones in a very controlled and deliberate effort:
if (version <= iPhone7) {
wait_ms(100);
}
For this specific case Apple's explanation, they started throttling down CPUs based on battery age to prevent unexpected phone shutdowns make sense. I do have issue with how this was initially messaged (a.k.a. it wasn't) but I understand why, consumer tech news tends to be very low information. It doesn't make sense for Apple to try to explain to users the details of every tweak to power management settings, and if they did it would either be lost on most consumers or be transformed in to negative headlines about how updating your phone makes it slower, in large part because the headline "Will the New iOS Update Slow Down your iPhone?" gets more clicks than "Apple makes Incremental Improvements to Power Management of iPhones as Part of Long-Term Strategy". So I understand the strategy of saying nothing.
Furthermore I expect most of the actual (not perceived) slow down is from dozens of conversations like happening:
Eng: New feature FooBar is awesome we should include it in the next iOS release. QA: FooBar isn't passing latency tests on our latest iPhone so you can't include it. Eng: Darn :(
<2 years later>
Eng: I still think Foobar is awesome, and look it very snappy on my current phone. QA: It works on our latests phones too, and users like it. Okay it is in! Eng: Yay.
This is fundamental to how software has been developed forever. We make faster hardware to be able to run more complex taxing software and than means older hardware will be slower in comparison.
It's not due to battery lifetime, it's due to older batteries' reduced ability to deliver high power quickly enough.
> …Apple wouldn't slow down the devices when they are plugged in.
No, because power is always pulled from the battery.
Of course I have no idea how iPhones are wired.
P.s the CPU and GPU scores both were noticeably lower (almost half) for otherwise perfect phone. Remember it's once benchmark score could still beat most Android flagships if not for the artificially introduced throttling by Apple..
At best it's a complete botch on Apple's part. I got that update and my phone was noticeably slower, right off the charger at 100%. My typical usage pattern has me plugging in my phone at night with usually more than 50% and rarely around 40%. I'm never in danger of unexpected shutdown.
The way things used to work, I'd drain an iOS device and somewhere around 20% I'd get a popup to switch to power saving mode, with a choice to do it or not. This is quite reasonable, usable, and the device remains fast until you're in the danger zone. Wanna tweak it? Move from 20% to 25%. Done. Worried that people won't switch to power saving mode and then have a shutdown? Make the switch mandatory at 20-25% or whatever.
But slowing down older phones all the time... it's hard not to be suspicious of motives.
https://www.geekbench.com/blog/2017/12/iphone-performance-an...
The thing to note here is that even in the charts of older devices, by far the largest peak is at max performance. In other words, most older phones, even running the new code, were not being slowed down when they were benchmarked.
There's no evidence I know of that every older iPhone is being slowed down, all the time.
Now, I certainly believe it's possible that your phone slowed down across the board after an iOS update, but there are a lot of reasons that could happen. See for example this story, where a customer went into an Apple Store on suspicion of a battery issue, but the performance problems were actually fixed by a backup and restore:
http://mglenn.com/blog/2017/12/22/apples-bungled-battery-fea...
But the decision process probably favoured shipping and trying to patch around it in software because form always wins over function.
Unsurprised by the lawsuits, sounds like a massive engineering fuckup to me.
A discount for a new battery is a laughable fix. It doesn't solve the issue which is going to reoccur once again, and it isn't free either. What did the customer do wrong in this case?
This has little to do with remaining charge. The batteries of the older phones could not, in certain cases, deliver the necessary power when it was demanded, which is what led to unexpected shutdowns. It was not directly correlated with your current charge.
"My household has both an iPhone 6 and 6s. Both have become noticeably slower, more battery consuming.
I just did the full backup and restore trick. We think it worked.
https://daringfireball.net/linked/2017/12/22/glenn-iphone-sl....
http://mglenn.com/blog/2017/12/22/apples-bungled-battery-fea....
My boss (she who must be obeyed) and I both "feel" that our phones are now more responsive and some features have resumed working like normal.
Such as retrieving voicemails, which had become unresponsive. For both of us. I had been blaming our cell provider (carrier). My bad."
Myself, I prefer my phone to stay on, even if that means it runs a little slower.
It's like buying a Tesla because of its incredible acceleration and smooth ride, but then in two years, that acceleration gets throttled to be similar to a Toyota Corolla to "maintain battery longevity." But hey, the new Tesla has even _more_ acceleration... It just seems like a dirty sales tactic.
The "slow-down" that no one talks about, which in my opinion is much more offensive, is that Apple creates new OSes that target a particular hardware platform and then push that same OS out to older handsets that don't have anywhere near enough horsepower to handle it. In particular I have an iPhone 4 running the last supported iOS. The device is effectively unusable. Everything is less than half as fast as it was out of the box. Gestures have to be done very slowly or the device misses them. Pressing the home button is not detected over 50% of the time.
I don't use this device having upgraded handsets twice, but it'd be nice if apple would put older iOS versions into maintenance and provide bug fixes sans feature updates so that customers aren't given the choice of buggy and insecure software vs. effectively unusable hardware.
I have an iMac from 2008 that is still working really well. I love what a fantastic machine they built, and for almost 10 years I could continue to upgrade the OS to get all the new features they developed.
With the "High Sierra" update I found out my computer was too old to receive the new OS. This made me really upset at first, because there are a few apps I want that I can never have on that machine.
But at some point if they continue to allow me to run the new OS it will result in an unusable device, like you are experiencing. But on either side of the fence you will have angry customers.
Apple can protect customers from updates that won't work correctly on their devices, or it can deny them access to new apps and features they want on old devices. Either way the customers will be disappointed.
Whatever you don't focus on, suffers. It's not intentional. It's a product of incentive structures around novelty. It's the natural inclination not to maintain the long view.
I don't expect the battery in my car to work as well as it did when it was new. If old enough the car won't even start.
I'm not looking at Apple here. It's everywhere. What seems to be missing is the continuation of your QA conversation that appears never to happen:
QA: "Foobar is truly painful on 3 year old hardware. You need to see if you can get it snappier, those handsets are still in support after all"
While we're about it we're also missing the QA conversation involving testing under poor and highly intermittent signal.
The more cores, GHz and mbs we get the worse it often seems to get.
Then perhaps Apple should have been transparent about what they were doing and told their customers who complained about the obvious slowdowns and were ignored by Apple. It was only until it was proven, without a doubt, that Apple was deliberately slowing down their CPU's did they finally admit to what they were doing.
Additionally, Apple has known about this issue for some time so they had ample time to correct the issue. Unfortunately, having the fastest mobile CPU on the market was more important than the side effects of a battery that could no longer provide the voltage required to hit those Geekbench scores.
The thinnest device is also an Apple thing. The decision to have a smaller battery to achieve that is also very obviously deliberate
"Never attribute to malice that which is adequately explained by stupidity."
Given upgrades are mandatory (user is spammed until he upgrades), given those features are planned without a flag to turn it off for older phones/iPads, and given it’s the only way to receive maintenance for the phone (i.e. security updates), it’s planned obsolescence.
Think every time you brought your car to a repairshop they would load it with unused equipment designed for bigger models, both clogging your trunk and making your car so heavy it loses 80% of its speed.
Wow, didn’t think the comparison would be so accurate.
Apple led the way on lack of repairability and expandability on their devices. Sadly other manufacturers slowly followed suit.
You likely need to turn off iMessage if you migrate to a non-Apple phone to get SMS or text messages someone sends you from an iPhone.
Locking certain iOS functionality away from 3rd parties - preventing browsers such as Chrome from using said functionality that Safari enjoys.
Want more?
Why they? Because that's their job - to help individuals coordinate at scale.
This is the most damning one. I recently (6 months ago?) switched to Android and I still have friends who cannot send me text messages. They're non-technical and unless I'm sitting in a room with them will not be able to walk them through the process. They have to do something, I can't do it on my end. So now I have a handful of friends with whom my only method of contact is FB messenger.
It's such a fucking scam.
I argue no. Apple has always been a head of the curve on removing hardware from their devices. This is upsetting if you use that device but it also frees up resources that can be used for other things. This includes space/weight/power wise in the device and economically factories making floppy/cd-rom drives now make something more useful. The problem with regulations is they are often slow and arbitrary. That is not what we want in tech.
I could imagine in 2010 that regulators (perhaps encouraged by a congress person that has a stake in CD-Rom factories), decide the iPad is a type of laptop and so much include a CD-ROM drive in it. Yes having every company make individual decisions is messy and inefficient but that doesn't mean regulations will end up producing better results.
Uh, such as what? The Nitro JavaScript engine is available for all apps now.
To avoid a case of 'no good deed goes unpunished', I do hope Samsung, Google and co get some attention over lack of updates for older phones/OSes too. Their approach of not updating perfectly good phones just leads to greater e-waste, not to mention wasting consumer dollars.
Has Samsung, Google et al provided any clarification on why particular phones do not receive updates?
Probably because they don't want to spend time supporting these devices?
It's more fair and conceivable that Apple should be forced to meet certain standards for updates they choose to release, than for any manufacturer to be forced to make updates for some arbitrary/indefinite period of time.
Security patches provide herd immunity and today it’s more important that everyone around you is patched than you or just you.
You share a metric ton of personal information with people and if their devices are compromised so is your data.
Now you might be on top of every vulnerability and wake up in be middle of the night to check up the NVD RSS feed but I would bet my life on the fact that the majority of your social circle doesn’t.
You wouldn’t accept Facebook or Google not patching their servers why would you accept your friends and family not patching their devices when they are in a much better position to compromise you if they are compromised?
Against my will, 99.5% of my email and social connections are saved by Google, Facebook, and the servers of whatever random apps slurped my friends' contacts. (Example: Venmo has a large and accurate friend suggestion list for me, somehow. I assume all those people have my number in their phone and their Venmo app matched it with my number.) So that ship has kind of sailed anyway. Their devices are compromised by design.
No real people hold this opinion. You don't hold this opinion.
You're saying that phone OS updates in general (whether from Google or Apple) are so low quality that its better NOT to get them on older phones?
I didn't take a general position; I expressed skepticism in GP's implicit argument that every update is a good update, and my frustration at the idea of having updates shoved down my throat. I included some other related ideas without explaining clearly enough, partially because I'm not sure how/if this should actually be regulated.
My old iPhone 5 started with iOS 6.0 and went all the way to iOS 10 but got noticeably less performant with each new major update. I'm certainly not arguing those updates were necessarily bad but I think it's worth discussing. Security and bug fixes are tethered with changes designed for much more powerful hardware.
A downright scummy company. /s
Of all the things Apple has been accused of over the years this seems like one of the most clear-cut examples where they didn’t do anything wrong.
You want to make political points? Go after them for iTunes, the App Store, something else. Somewhere where you have a leg to stand on.
“Your honor, they tried to cause the phones of the American people to be more reliable! I demand justice!”
Funny thing is, I've owned non-Apple phones for years and I've never had any "randomly shut off." If you pay half attention to the battery capacity icon and/or warnings the devices gives you, any 'shutting off' due to no more battery charge will not be random.
Maybe Ford should recall their cars because the tires eventually wear out?
I always attributed it to an unintended side effect of the case itself being used as a heat sink. Battery gets very cold very fast and loses peak voltage. Would be interested to hear if there's a more accurate explanation.
The voltage written on a standard battery cell is the "average" voltage the cell will produce across its life time.
As amperage is spent(?), the voltage will drop slightly, so the cell will start out higher than listed and end lower than listed.
What is likely happening is that the when a certain app is started, various parts of the hardware is clocked up to handle the work load. This result in a higher draw, and a sharp drop in the measured voltage, in turn the power management logic considers the battery nearly drained and shut down.
Never mind that unlike older battery chemistry, lithium batteries do not take kindly to being fully drained. Thus there will be a programmed safety margin involved.
eh?
My PC did get hacked in the 90s when I was running something like kernel 2.0.36.
Look at all the trouble people are having with the meltdown/spectre updates. I haven't updated, I have no trouble. The odds of trouble if you update I think are much higher than the odds of any trouble from not updating.
Keep in mind I'm talking about a PC behind a firewall not an internet facing server.
I cannot get behind the criticism of Apple on this specific issue. I think they were doing the right thing and I think adding a software option to allow the user to disable the governor algorithm (for lack of a better term) is akin to giving the user enough rope to hang themselves with.
It happened while I was travelling, and having the phone shut down isn't helpful when you're trying to navigate via Google Maps, or take photographs, or while you're trying to hail an Uber. I would have loved to have the option to slow my phone down instead in that circumstance instead of just random shutdown and reboot.
In and of itself? Nope. But someone should be asking Apple why they have 3, if not 4, generations of phone now that has to have this software "fix" in order to continue to operate. Looking at forums, some people's phones are getting throttled well before Apple's 80% battery capacity mark. Someone needs to be asking why. Some people's phones are getting throttled after only a year. Why? Someone needs to ask.
Stepping back a bit, I don’t recall hordes of reports of random iPhone shutdowns in early 2017, but it certainly did happen some of the time. This update slowed down every iPhone with an old battery regardless of if that user was ever really likely to experience random shutdowns. The fact that Apple obscured this change also makes it a bit more suspect.
> This power management works by looking at a combination of the device temperature, battery state of charge, and battery impedance. Only if these variables require it, iOS will dynamically manage the maximum performance of some system components [0]
An especially cold winter.
Could just be timing/environmental compared to when the original change was released.
- stop powering the device when the charge gets into the zone where sudden dropouts might happen (i.e. lose about 5% capacity)
- accept sudden dropouts and have users experience devices that suddenly fail (i.e. lose about 5% of their useful charge since you can no longer depend on the device being on)
If you're inclined to think the worst of them, well, obviously this reinforces the idea that they're greedy, anti-consumer schmucks who deserve to be taken down a notch.
In my book, that's the best news to come out of Apple in years.
This is a phone that costs with sales taxes, over $1000. At one point, it was so slow that I felt like throwing it on the wall, and breaking it into a thousand small pieces. (There even was a lag between me typing on the on-screen keyboard and those characters appearing.) I replaced the battery (about a year ago), and the speed improved a bit, but not completely.
I had to get a new phone because it was (still was) simply just too slow. Honestly, there needs to be some sort of class-action lawsuit and compensation for iPhone owners who have literally been swindled by this.
What new phone did you end up getting? I hope it wasn't another iPhone.
But I purchased the iPhone X with AppleCare+ about a month ago. It was a huge purchase. The phone was $1,149, AppleCare+ was $199, and NYC sales tax was close to $120. In total, I spent $1,468.
Unfortunate, but software is hard.
You do realize that the update that "enabled" this feature was 10.2.1, which came out in January 2017? Any slowdown you saw beforehand was unrelated to this issue.