Hacker News Reader

Comment by rekoil | Hacker News Reader

> by reducing performance like Apple does with old iPhones

What Apple did was clock down the CPU of iPhones with degraded batteries in order to reduce peak load which would otherwise result in crashes that would have made the device borderline unusable below some charge percentage.

Yes, they should have shipped the feature off by default and with a toggle, but the feature itself was amazing as it let already old phones last longer than they otherwise would have.

Apple deserves a lot of shit for their bullshit surrounding replacement parts, but this particular topic is something they get an undeserved amount of shit for in my opinion.

View on HN

> "What Apple did was clock down the CPU of iPhones with degraded batteries in order to reduce peak load"

It wasn't what Apple was doing that was the problem, it was the fact that they were initially quite secretive about it! People just started noticing their phones slowing down without explanation. Now days iOS tells if your battery no longer supports peak performance, which is the right thing to do.

> "Yes, they should have shipped the feature off by default and with a toggle"

Oh, it should always have been on by default. Most people would much rather their phone run a little slower than risk random crashes and shutdowns! But iOS should have told you about this from the start.

I wonder how many products have components that downgrade under specific situations and they never explicitly tell the customer - and most customers don’t ask because it’s “complicated tech” they don’t want to know about it.

I think a lot of laptops do this because their cooling system is too weak for the CPU.

The Lenovo Thinkpad t480s comes to my mind: with factory settings, the CPU was unable to reach its advertised boost frequency nor maintain its max non-boost frequency for more than a few minutes. After 5 to 10 minutes, the machine was usually running at 300/400mhz under its advertised clock speed.

Undervolting helped a lot but could not fix the issue completely, under clocking was the only solution to maintain a consistent clock speed.

IMO, at least with x86, there is a point where it simply doesn’t make sense to cram higher clocks into a laptop. The last Intel MBP could be had with an i9, and IME if you left turbo enabled, even at idle the fans were annoyingly loud. If you pushed it, you’d get a quiet jet engine as they struggled to keep temps just below 95 C.

It was usually less annoying for me to just disable turbo, and only turn it on if I really needed the boost.

Thinkpads, for a long time have had a fan speed that's possible, but never reached by the firmware's automatic mode. On Linux, this sets the fan to its true maximum:

    # echo disengaged > /proc/acpi/ibm/fan

I use the thinkfan utility to run the fan more aggressively, including 'disengaged' over 70C.

How's your hearing? If any left.

I put in foam-tipped earbuds if I'm doing a big compile or something, but the Ryzen 6850u runs pretty cool most of the time.

Indeed, and for the t480s we could undervolt the CPU to help with thermals but these are not normally accessible to the end user.

So if you are an average user of a Lenovo laptop without technical skills, you probably have a machine that does not meet the advertised performance, and no way to fix it.

This is likely the case with all but gaming laptops. Most users won't care about performance under consistent, heavy workload, as much as it loading that .docx in 2 instead of 6 seconds. That burst of performance is what makes the system feel snappy and fast. If the user is happy, it was well worth the money too.

plenty. But normally the degradation is more gradual and automatic. For example, as the innards of your clothes dryer gets clogged up with lint, at some point the thermal cutout will cut in and out to briefly reduce the heat.

End result is your old clothes dryer starts taking longer and longer to dry clothes.

A lot of laptops do this when running on battery power. Do they pop up notices telling users they are running in reduced power mode?

I recall mine sometimes doing so if I unplug while it is on. Never occurs while it is booted on battery.

Laptops and phones also notoriously slow down to manage heat, and that is also never told to the user.

Clog the heatsink on just about any PC and it'll happily and silently thermal throttle itself.

That stretches the definition of silent. On MacOS you need command line utilities or third party software to observe throttling, but on Windows this information is easily viewable in the performance tab of the task manager, and Linux exposes this information in /proc.

Being buried away in technical debugging tools is 'silent' for 99.9% of users.

A common task that repair shops do to "fix a slow PC" is clean dust from the fans and heatsinks to reduce throttling. Users don't know or understand that this is happening, they just know their system is slow.

Maybe this is just me but I always assumed USB cables had indefinite lifespan until I happened to read a datasheet.

>People just started noticing their phones slowing down without explanation.

Also, Apple already had somewhat of a reputation; this was far from the first time they would do shitty things, blame the users, and/or refuse to communicate what's going on.

It's a classic case of 100% right and 100% fired: https://www.youtube.com/watch?v=nDO24U3hMkU

Well, duh. You're holding it wrong. /s

I never could get that to not work. No matter what I did the signal didn't degrade.

because by the time you tested they had already boosted signal power beyond all guidelines

Thin hands and hollow bones?

iPhones with degraded batteries have the tendency to shutdown on peak loads. Default is fine but only after a big fat pop-up with the option to deactivate it.

As if the typical user (not anyone that browsers hn) actually reads the _big fat popup_ and would just click through anyways. With no care or worry about what that pop up said or the implications.

Apple “You clicked a pop up that said you accepted random shutdowns and reboots due to a degraded battery instead of allowing us to degrade performance to enable a stable experience with a degraded battery.”

How would you explain the above to tech illiterate people? Increase suppose calls because you allowed a user to chose between degraded yet stable performance versus seemly random (to the user) reboots and shutdowns?

Because one looks like a hardware defect and the other like a problem of old age and is "easier" to accept for most people.

Default to speed and then the first time a device has crashed "your battery has worn to a level that we will have to start slowing down your device or it will keep crashing, will you accept these crashes or slow your device?"

That would be transparent and also allow decisions, if AppleCare also allowed for free battery replacement after receiving that message it would be a PR boon, not a kerfuffle.

Maybe.

But you might overestimate technical understanding of people.

"Letting people know" comes with it's own challenges.

There are a lot of choices an OS makes without letting it's users know.

It would be quite overwhelming to communicate all. And in making choice what to communicate mistakes might be made.

Screaming disgrace is a bit exaggerated.

When this came out, most people were distrustful because it was a secret. It was seen as a way to force device upgrades, not save devices. If they had been transparent, even in documentation, (and allowed opt out) it would have curbed some of this backlash.

It may well have, but that doesn't mean the backlash was reasonable.

I believe it was perfectly reasonable based on the information at hand. Apple gave no reason to believe otherwise and behaved in a way that appeared deceptive.

I was an iPhone 6 user at the time and an Apple fan. I think the secrecy surrounding the move means that the backlash was reasonable.

Consider: a pop up saying "Your battery need replacing, it is operating in slow mode to avoid crashes" v people thinking "Damn, my phone is slow - I better buy a new phone"

Some devices had a small performance hit, but some had had their clock speeds to less than half, making devices laggy even during normal usage.

By itself this is already a problem, but they also decided to not allow you to disable the feature or reverse to an older iOS version. I don't think they informed their Apple store employees either, at least I was told to buy a new phone. They had to be sued first to add a simple setting, ffs!

So yeah, I think the criticism is valid and deserved. I'm sure they had good intentions, but the implementation was shit and not only they created issues for some iPhone users, but also profited with new iPhone sales (not me, that was my last iPhone).

They didn’t explain what it did. The feature was amazing, the presentation misleading.

No, Samsung and others spread a bunch of FUD that Apple was crippling their phones to force people to upgrade, and Reddit and HN apple-haters ate it up.

Still to this day people on HN repeat it like it's fact, and when you correct them, shift the goalposts to "well they didn't TELL people."

This is not a goalpost shift. It's THE argument. I'm not sure what arguments are there for alternative narratives.

Do we know they actually didn’t tell people? Does anyone read through the “What’s New” pop-ups after they upgrade the OS, or just blindly dismiss them?

Apple provides a very comprehensive user manual that ships as a pre-installed app on every phone. There are large headings describing the latest features right when you open it.

I don’t recall being unaware of this feature before it became a “scandal”, but I’m one of those weirdos that reads the release pop-ups.

Let's say you read that "what's new" pop-up and still update your phone because you expect it to work like before, but it doesn't because you're one of the lucky ones that had the clock speeds cut in half.

What do you do now? Your phone is laggy and there's no setting to disable the feature or a way to revert to an old iOS version. You go to the store and they don't tell you anything about a battery replacement, but instead suggest a new iPhone!

The feature was useful, the implementation wasn't. The issues weren't caused by the user not reading the changelog.

> they don't tell you anything about a battery replacement, but instead suggest a new iPhone!

You keep repeating this, but battery replacements were a thing before this whole debacle. I got them a few times, I think they were like $79 (yes definitely more money than the sus extended batteries I'd buy off eBay for my HTC Evo). I realize society collectively forgot about them once the back stopped coming off the phone, but surely a tech savvy hackernews would realize that they couldn't use the same battery forever without replacement.

To replace the battery you need to know there's a problem with the battery. But the iPhone didn't tell me the issue was the battery. It happened after a software update if I remember correctly, so I did everything from formatting it to reducing the graphics load by using the accessibility settings... nothing worked. When I went to the nearest Apple store, the guy told me that I should probably get a new one.

I don't blame him for recommending a new phone, after all apparently Apple didn't inform all their stores about the change, so they probably had no idea that the fix was a cheap battery replacement.

I think it's equally possible Apple actually didn't know how many phones would be hit by the feature when they launched it.

They knew that old devices (especially common with the well-selling iPhone 6) would fail at peak demand, and wanted to address that, but they probably didn't have an accurate figure of exactly how many devices would be covered by the feature as initially configured.

That it took them as long as it did to react to the resulting situation is unfortunate, but could probably be explained by corporate culture at Apple, they don't generally like to admit mistakes (see the Trash Can Mac Pro press meeting a few years back).

In the end it was just a huge PR fuckup, if they had properly considered how many devices would be affected they would have realised they needed to communicate more clearly, they could have avoided lawsuits and honestly if they had played their cards right from the get go they would probably have received praise instead of backlash, but instead here we are 7 years later still explaining how battery chemistry affects integrated circuits on technical forums.

There was a communication problem, but the main problem was the slow iPhones. I bought the phone and Apple then made it slow and laggy. I wouldn't be here complaining if they added a setting to disable the throttling (which I probably didn't yet need at the time as the phone had never shutdown before the update).

> Do we know they actually didn’t tell people? Does anyone read through the “What’s New” pop-ups after they upgrade the OS

I'm an avid reader of pop-ups and release notes. They didn't tell people.

> It's THE argument

From people I've talked to, it seems like the VAST majority of people are under the impression that Apple does this as a planned obsolescence practice.

In fact, this thread is the first one I've come across where most people seem to get the point of the feature

The planned obsolescence was when Apple put a small battery in a flagship phone that could no longer produce the required voltage and/or current to keep up with hardware demands despite being only moderately used. Apple is the same company that brags about forever years of software updates, but plenty of people were running into this issue after only a few years, including my own mother who's iPhone couldn't even make a single phone call despite the battery reporting 90% state of charge.

Notably, this will no longer be an issue on modern iPhones because they have much more reasonable batteries.

Apple underspecced the batteries in iPhones and then hid the software kludge they made to hide that fact. That's why the reaction is warranted.

This is a tangent to the original point, but I'll play. Apple iPhones have long had some of the highest battery lives of smartphones on the market despite having some of the smallest batteries. Since they only have to worry about a single OS, they can optimize away to achieve this. Smaller batteries means less environmental harm, lighter phones, and more space for other parts. There are endless reasons to bag on Apple, but the fact that they can achieve much better battery life with much smaller batteries is not one of them

No, Apple's throttling to prevent crashes is no more "planned obsolescence" than how Samsung (or any) laptops thermal throttle when the heat sinks get dirty.

Designing a product to fail gracefully is good engineering practice.

Apple made the right decision to make the phone go slower instead of crashing. Somewhere out there, there's someone making a 911 call on an iPhone with an old battery.

The fix had an impact on a core metric of the device: performance. This was the best technical solution to gain longevity, and it was in the interest of users. But doing it under the radar left everyone wondering if it had ulterior motives and let the rumor mill go crazy. Transparency is good even when everything is upsides & roses, and especially important when there are some compromises to be made.

Very well put.

My iPhone was affected by this. Let me tell you my experience, without "Samsung FUD":

- The phone became slow, laggy animations, etc.

- I had no idea what was causing the problem.

- There was no way to return to an older iOS version or a way to disable the new behaviour.

- I went to an Apple store and was told that I probably should get a new phone.

I'm sure they had some good intentions, but in practice they degraded my experience too much and profited from new iPhone sales that didn't need to happen. Apple had to be sued first to add a setting to disable this feature and start offering cheap battery replacements (by then I had already moved on).

From the implementation of this feature to the way they handled it, it was all Apple. Samsung had nothing to do with my bad experience.

Not everyone was affected this much by the change, but some where. And that's a problem.

Except they only HAD to do that because the battery simply did not have the capability to continue providing enough voltage and/or current by the time it got to about 20% degraded, which for many people could happen in just a few years. This only happened because they were putting stupidly small batteries in their phones to shave another millimeter off of the size that nobody cared about.

Either Apple did the proper QA of the battery to understand how it would behave as it got to end of life and knowingly put a battery into a $800 phone that wouldn't be functional before the phone would run out of software updates, or they FAILED to do proper longevity testing on the battery of a $800 phone.

It wasn't planned obsolescence when they added that code to keep the older phones working. It WAS planned obsolescence when they put a pitiful and underspecced battery into their $800 phone.

Apple's know-how in battery management is always a couple of generations ahead of its closest competitor.

My 2008 MacBook Pro was able to detect battery problems one year before they started to manifest physically. They did it by adding another controller and an independent firmware to the battery itself, but there was no better tech at that time.

M powered MacBooks added auto hold-off (discharge to 80%, keep it there) or postpone charge (because the battery is used too rarely), to the list of tricks they know, and these are just the visible ones to the user.

> discharge to 80%, keep it there

Where? AFAIK this is a feature you need a third party app like Aldente for. The OS does not offer any way to control battery charging level. Yes, it tries to be smart and guess when you need full charge and sometimes decides to charge to only 80% but that doesnt work for all users except the ones with very regular routines and there is no manual control.

I meant the automatic one. My charge cycles are not very predictable, but periodically my Mac says that It'll hold-off charge, and uses battery until 80%, then uses power adapter. Then charges the battery fully later in the day if I'm connected.

Sometimes it flat out refuses the charge. Then I disconnect and use the battery until 30% or so. Then charge when convenient (clarification: It'd possibly charge if I hit a certain percentage, but I always disconnected voluntarily, to give a good use to battery. I lasts a day. Why not?).

My MacBook Air M1 has 165 cycles in 3 years, at 90% of design capacity. I think this is a very solid management of the battery for a computer used every day literally.

My MacBook Pro is at ~85% after 8 years. I don't remember total cycles now.

> Why not?

I find running on battery, lots of things are throttled or disabled. CPU is slower for example. That's probably a good reason for why not.

I generally do mundane tasks like SSH sessions or zoom meetings when I'm on battery. Heavier tasks are always offloaded to my desktop system, most of the times, so running on battery here and there doesn't impact many things in my case.

A video call though isn't that mundane. Google Meet often has my cpu and fans spinning. I do prefer Zoom as it uses fewer resources. But of course, it will work on battery.

For an M1 powered MacBook Air, none of them creates a significant power spike on the system. I use both regularly, generally one hour at a time, and battery percentage doesn't take a noticeable hit after the call, and the computer doesn't warm up.

>I find running on battery, lots of things are throttled or disabled. CPU is slower for example.

FWIW, the performance and CPU features are the same on Mx based MacBooks whether they're plugged in or not.

> Apple's know-how in battery management is always a couple of generations ahead of its closest competitor.

> M powered MacBooks added (...)

Things that Lenovo laptops have had for a generation?

IBM Thinkpads have "cap charge at n%" setting since 1999 or something, but it's not smart. You set it.

It was also present in HP and Dell's business notebooks for quite some time, but again, it's not smart. Just a pre-defined percentage.

Apple manages maximum charge capacity automatically depending on the way you use your computer.

I've had smart charging in other laptops for ages as well. I'm not saying Apple isn't doing good, I'm just disputing the "Apple is on another league" sentiment, when it's mostly based on a lack of knowledge of what's happening outside the Apple ecosystem.

Well, I didn't witness on any other current generation device, at least yet. I'm not a fanboy. I'll be happy if others are doing this in some form, esp. if it's independent of the OS itself.

My colleague just had to replace his HP Elitebook Notebook because the battery had bulged so much it had cracked open the case.

For this to be happening in 2023 means that some manufacturers at least, seem to be not managing this potential problem at all.

N=1 doesn't really tell you much of anything about anything.

Well, my wife's work supplied EliteBook's battery has been replaced today due to excessive swallowing, and it's definitely younger than my 2014 MacBook Pro.

An entire production run of Macbook Pro 2015s had to be recalled and replaced because the batteries were inflating and bursting into flames.

...and Samsung recalled two series of their flagship phones to replace their batteries, and fail (and cancel the model altogether). Because of their failure, planes started and still continue to announce phone charging related security measures during take-off.

Every company makes big mistakes. Also Apple once built laptops with self-cooking GPUs, a couple ones very fragile keyboards, a series with too-sensitive screen coatings.

We can write tons of problems with every producer. The most important metric is whether they're learning from that or not. As far as I can see Samsung and Apple learnt their lessons.

> Apple manages maximum charge capacity automatically depending on the way you use your computer.

Apple cannot predict where you are going to use laptop on AC power at home or whether you are going somewhere and need a full charge. So manual control is better.

There are apps for that if you want, and if it enables hold-off, you can always say "No, I need full charge now, so please charge".

That’s the weird part, they actually can. It’s supposedly one of the machine learning features, where it learns your schedule and makes daily accurate guesses of by when you’ll unplug and plug it back in.

> ... they actually can.

Yes, it does. Even my 7 years old iPhone X does it. I use some features at very specific times, and it suggests me "do you want to do X?" around the time I do that. So the models work well.

Also, the phone generally enables optimized charging if I put it to charge too early. Scheduling to finish a couple of hours before I wake.

> I use some features at very specific times, and it suggests me "do you want to do X?" around the time I do that.

You're conceding that you have a predictable schedule. Folks on the other side of this (who believe it can't work nearly as well as manual control) don't have a predictable schedule. Think of it like {hybrid workers who use a desk all day the same 3 days a week, and are nomads the other 2 days a week, and the ML system can figure this out} vs {on-site workers who use a desk all day 5 days a week, but are nomads a few random days a year, which always catches the ML system off guard}.

Both are valid. A setting that caps it at 80% unless you hit a button for "I'm about to spend a lot more time unplugged than usual, so prepare the battery for that" is the only way to handle unpredictable circumstances. Just like preparing an EV for a road trip.

As said by me and others elsewhere on this thread, this is of course a valid requirement, and tools have been developed for permanently capping charging percentage on Macs for quite some time. I don't have an objection to that.

My Macs charge routine is way more hectic than my phone's which is very predictable, and both of them guesses what I'm trying to do correctly 99% of the time, and that's fine by me.

What I said is, it's working fine for me, but the system is open to tuning both out of the box (I need this now, please charge), or with some small tools (it was called Aldente IIRC), for differing, equally valid needs.

I don't think we're in disagreement here.

BTW, your nick reads ********. Is it hunter2 or something? ;)

It’s not new in Apple devices, either…

> My 2008 MacBook Pro was able to detect battery problems one year before they started to manifest physically.

That may or may not be noteworthy. Despite half my family using MacBooks, and some having had major battery problems, I’ve never heard of such a notification as you describe (including at least one where I’d certainly have expected to hear about it). And the similar system SMART for storage devices is notoriously unreliable: I’ve personally had devices with dire warnings chug along merrily for plenty more years, and observed one or two failures that never received any warnings.

I used three batteries on that particular MacBook and it always shown a "! Service Battery" warning before things went bad. The battery life, and shape of the battery was indicative of anything, either.

Oh, don't bring up SMART. I remember long-testing >100 disks, and most of them passed (it took hours), and all of them failed within 10 days after facing real workloads. We needed a temporary storage for migrating things, but we lost no data, because it was just a copy operation.

All the shit was deserved for altering phones without the consent of the consumer. I don't care how good they had in their hearts, get the user's consent first.

Reply on news.ycombinator.com