In case anyone missed it, the author of the submission added an update [1] saying that after some time the issue went away:
> Aaaand it's fixed... I didn't do anything specific. It took a few hours (?!) after the upgrade. It scored many times low, but now it seems to score at a normal level (3000+). Really wonder what the explanation for all of that is.
I'd bet it is, though i don't have any insight - someone with the necessary knowledge and a jailbreak probably could take a look at the process list and know for sure...
Speaking of slowness - not sure if i'm the only one, however Spotlight-searches on my iPhone (8 plus) have become dog-slow since sometime iOS 12 or so - before, the results would be instant - now it takes something between 3 and 5 seconds to return a list of Apps for example :(
Something similar happened with the macbooks a few months ago. Apple changed something in one of their services which made it impossible to open non-Apple applications and made the entire computer grind to an halt. After thirty minutes everything just started to work again for no apparent reason. Turns out that their service for validating application certificates went down for some time.
Not sure if it’s related at all, but it’s still sad to see that our devices have become so dependent on phoning home that they can become unusable when the home is not picking up or answering incorrectly.
I updated this morning and continued working on an iOS app. Everything felt slow. I thought I did something stupid in my code. Apparently not. As per this Geekbench 5 test, the phone scores ~1200 where it used to regularly score ~3000+ (!!!).
There's a lot of rumors what's the cause of this but no official explanation from Apple as of right now that I can find.
Every time there’s an iOS update, the internet goes crazy about their phone being slow or hot or not holding a charge. Then, a day or two later, the background system processes finish doing their post-upgrade homework and everything goes back to normal.
While this might be appreciated by a small number technical people it would just be confusing to the overwhelming majority. The cell, wifi, and battery indicators all provide information the user can affect by moving the device or plugging it in. An indicator would not provide any actionable information to the user. Also progress bars and time estimates for tasks are typically expected but in practice very difficult to make accurate. So you could not even accurately tell the user how long they need to wait for the device to finish it's background work
I don't see why they shouldn't include this when they include the info when it happens for a mac. When you get a new mac or it updates, you can see that spotlight is busy reindexing which explains the slowdown. I'm surprised there is not some one sentance explanation somewhere in the phone, like "processes will continue to update in the background and phone performance may be affected until this finishes" and boom, the OP article would have never been written.
"Background system processes running" wouldn't be obvious to non-technical users. It might even be distressing. If the process isn't something that can easily be explained to a regular user, I can see why Apple wouldn't try and communicate about it.
Though I could see maybe detecting that a user is a developer and showing a message only for those users
But unlike this (which is timed in hours) in Android it's done in a minute and then the notification is gone. Maybe Apple should start by doing the update in minutes first.
I find it really bizarre that it takes hours to run whatever background process iOS has after an upgrade, especially because in Android's case this is AFAIK rebuilding the Dalvik bytecode while iOS should use a compiled language (ObjectiveC/Swift).
Does someone knows what is iOS doing at the background?
Hmm, if I specifically go looking for reports of benchmarks going down, it does seem to be a much more prevalent story with this version than any previous one.
Reports of post-upgrade slowdown are pretty common (google it for various versions), but this is the first time I’m seeing lots of reports of post-upgrade benchmark hits.
Did you know that batteries are made of chemicals and don’t last forever? Over time their performance degrades, but you can use them for longer if you don’t use as much peak power.
If your update system requires 1-2 days of "post-upgrade homeswork" time and slows your phone by 50% in the process, you've designed a very poor update system.
Apple successfully upgrades a huge majority of all their phones in the field, and nearly all of those users are barely even aware of it. They plug their phone in at night, sometimes in the morning there's a message about a new version. Maybe they notice their phone is slow for a day or two, likely they don't. No major incidents involving mass-bricking of phones, no rollbacks, no data loss... arguably the most successful update system in the history of technology.
> It’s just true that data structures change over time
Sure, but it seems only 1 of the OS's takes this long to be fully usable post-update. Why is that? Does Android not ever have to change data structures and if so, why doesn't it take their OS 24-48 hours of post-processing to finish the update? You find this acceptable?
Android doesn’t have the same feature set. Comparing the two like this is an obvious fallacy. For example, Apple does their ML e.g. for photos on device for privacy reasons, whereas Google does it in the cloud.
I’m not saying ML is the cause although it could be. I’m saying they are different products with different values, so expecting them to behave the same doesn’t make sense.
And yes, some background processing after an update is more than acceptable. It’s desirable if it means sending less data to Google.
Aaaand it's fixed... I didn't do anything specific. It took a few hours (?!) after the upgrade. It scored many times low, but now it seems to score at a normal level (3000+).
Really wonder what the explanation for all of that is.
Background process doing some sort of migration or other data conversion between versions? Strange the device wouldn't wait until during the night while plugged in, but without context, that might not be an option.
Kind of sad that we don't have a top or process viewer. Or free -m / open system. This makes me yearn for an open source OS on my phone again, I may have to look into a pine phone.
If you connect your iPhone to your Mac you can run Instruments on it to figure out what is going on. I used to recently to figure out that Slack had been making my entire phone slow because it was busy looping in the background.
I agree, but it shouldn't be a classic process viewer. It should be a task viewer which has some context or understandable description. Like "photo analysis job" or "iOS 14.5.1 data migration job".
Yes... I remember back in the day I had terminal with root access on a jailbroken iPhone. You could install almost any package on your phone via terminal.
One of iOS' more annoying quirks is its opacity and lack of user affordances.
The volume buttons don't seem to affect haptic noises, but the mute button does. The OS locks up for about a minute when you change passcodes while it re-crypts everything, but from the user's perspective it just looks frozen. The implications of opting into their 'find my...' and iCloud networks are not adequately explained. The devices take 15-30 seconds to actually cut power after they appear to shut down, and during that time they won't respond to any button presses.
The list goes on, using a modern iPhone feels like you are constantly negotiating with a disingenuous UI. The OS regularly claims to be done with something while it continues working furiously in the background, and it is full of undocumented exceptions to common settings.
Is there mention of this in any of the dialogs that one is asked to click "I have read and understand" that were not actually read? I remember Photos telling me that it was scanning the library to do their new layout after whatever update that was. That was to that specific app, and not a system wide notification, but it also said it would be paused until it was plugged in.
I could see a new something in the Settings that indicates post-OS upgrade showing migration status being useful.
If there was I didn't see it. But that is no surprise. Its not a way for them to escape a notification either imo. Especially if its slowing everything down.
Yeah pretty sure it’s this. Have seen various complaints about both macOS and iOS sluggish performance post upgrade to end up being just temporary while caches / indexes rebuilt, file systems migrated etc
Likely, yes. The thing is, the phone does not "heat up". Perhaps they throttle the whole phone down globally and let the background update processes run at a slow speed/low priority so they don't heat up the phone too much.
My FaceID stopped working. I cannot %100 relate it to the update itself, but after the update, I also enabled the unlock with watch, after one day, I decided it it not for me and disabled it. After that my FaceID entirely stopped working. I have removed face id to add it again, but cannot add it back, it keeps saying "move iphone a little lower". Tried everything short of resetting to factory defaults and nothing has worked yet.
Yeah I have 14.5.1 for a few days now and my 11 pro max scored 3509 which is pretty much bang on when it was new. I just went and checked. It’s actually gone up since it was new. I have 3 benchmarks in the 3400 range so if this is an issue it’s either temporary along the lines of someone needing to let background processes finish or some phones are affected but not others.
It wasn’t artificial, it was required to prevent random shutdowns due to over-voltage on a degraded battery. The fine was due to them not telling the consumer like they do now in ‘battery health’.
I never experienced a random shutdown on my old iPhone despite not allowing the software update for years. Finally, when I did update the phone, everything was noticeably slower, instantly, so that sounds more like good PR management than actual consumer-minded thinking to me. Yes, devices get old and batteries get old, but my understanding is that this happens even if you replace the battery with a new one, which would negate this argument.
> Aaaand it's fixed... I didn't do anything specific. It took a few hours (?!) after the upgrade. It scored many times low, but now it seems to score at a normal level (3000+). Really wonder what the explanation for all of that is.
[1]: https://news.ycombinator.com/item?id=27131683 .