I find myself pushing off macOS updates far more often than I do Windows updates. They seem to take far longer and I feel like it's a black box as to what's really going. If I have it hooked up to an external monitor, there's a chance it just stops displaying the progress bar for the update on the monitor for periods of time, making the entire process seem like a black box. And then once it comes back up, sometimes I have to open the lid and login with the built-in keyboard just to resume using everything connected to my dock.
I've been wondering why MacOS updates take so long. Surely with how MacOS has a read only root volume and APFS can do snapshotting they can build the new boot image while the system is running and then just ask for the user to reboot into the new volume, after that consolidating the snapshots by deleting the old boot image.
One potential source of the long update installation time is going to be in how they ship firmware updates as part of their OS updates now and have done for a while (Note that they stopped shipping external firmware update packages for years).
On Mojave, there is an option to disable automatic macOS upgrade checks (not sure about Catalina/Big Sur). I only get nagged for security updates, though I believe you can disable those as well.
It's incredibly frustrating when a software update offers no feedback. I understand Apple doesn't want to burden its non-technical users with the details of what the update is doing but it gives them no recourse when something happens like as described in the blog post.
There should be a hot key that can display the log of the software update process. That way the user can search for "MacOS update stuck on reticulating splines" to try to find possible solutions. I really don't understand why Apple makes it so hard to diagnose issues with their systems.
And all the programs that stops working with each update. Have to hold off just to ensure that all programs are updated. Hated that each update broke something in my workflow, one of the reasons I moved away recently.
You could reasonably refer to running in clamshell mode as headless, but calling a system with external monitors connected headless does not make sense.
The only thing in this blog post I can identify with is the weariness of the Apple update requirement treadmill. If your phone updates to the latest OS, you have to use the latest XCode to build or test an app on it, to use the latest XCode, you have to be on the latest mac OS, to be on the latest mac OS you have to have supported hardware or you can hope that someone has come up with a hacked patcher to get you on the latest OS.
I hate this as a developer. I feel this is done to force you to buy newer hardware sooner. The SDKs could be decoupled from the Xcode version.
You can work around it a little bit by installing the iOS device support files manually without updating Xcode, but it only allows you to deploy to the device. If you want to build against the newer SDK, you have to update.
Going forward this is one of the reasons I don’t want to personally develop any more apps, as I don’t want to be forced to buy new macs every five years or so.
Not to mention with every new release of Xcode, I seem to be immediately unable to build any (or at least many) of my previous projects without much manual tinkering.
Jumping through hoops to keep (or retrieve) older SDKs etc.
Admittedly I haven't tried this in a year or so, but I imagine the experience hasn't improved.
In no other ecosystem have I found existing projects to break so quickly and reliably with every update to my IDE. I would dread Xcode updates for this reason.
As someone who has used Apple computers for decades — including the ][+ and dozens of Macs — I feel I must rebut some of the specious arguments being made here. To wit:
* Mac software quality has indeed declined precipitously over the last decade.
* It is not “Apple bashing” to point out that fact.
* OP never claims the failed upgrade is the only example of said quality decline, but rather that it is one of many representative examples of that trend.
* One failed upgrade in 20 years is not good, nor is it anything to brag about. As the OP pointed out, what is a normal computer user supposed to do in this scenario? This simply should not happen, ever. Features should be dropped, deadlines should be eliminated, and work on everything else at Apple should stop until they can ship an upgrade that does not fail, or at least fails gracefully with an easy way to recover.
* For those who point out they upgraded just fine: “It works for me” is not a compelling argument. On the contrary, it is a terrible argument that should never be made, or be taken seriously, by anyone, ever.
In past years I always looked forward to upgrading to Apple’s latest and greatest. Now I dread upgrades and avoid them until I can’t avoid them any longer. I am still running Mojave, and experiences like the OP’s only reinforces my confidence in the validity of that strategy.
>It is not “Apple bashing” to point out that fact.
It is Apple bashing when you make dumb extrapolations out of a 1 in 19 year event that happens to 1 Apple computer out of how many millions (billions ?) that Apple have sold over the same 19 year period.
> what is a normal computer user supposed to do in this scenario? This simply should not happen, ever.
Failure of a software update should "not happen, ever" ?!?!!?!?!?!?
Have you ever worked in a technical role in IT (or any sort of technology) ?
As I already explained, there is no extrapolating: pointing out one example of poor software quality does not mean that is the only evidence at hand. In fact, you are the only person who seems to be making that argument. Which is nonsense, since this incident report is one of many, many documented problems that clearly indicate a decline in software quality.
I have worked in a technical role for many years, and if my department shipped Big Sur, I would find immediate replacements for everyone in that department.
I have a MacBook air with a 128GB SSD (I was not involved in this purchase decision :). Upgrade to Big Sur would fail repeatedly without any error message. It would do all the rituals (including gigabytes of downloading), and pretend to reboot. After the reboot, nothing has changed!
It was a good 3-4 tries later I thought to free up the disk a bit and then the upgrade went through.
I think they can do a better job of surfacing error messages to the user. Insufficient disk space is a basic situation I would expect them to have a proper error for.
My work bought everyone the 128 models too. It’s barely usable as a machine and very difficult to keep more than 5GB free, I know most Mac advocates will never know this because they don’t buy base models but anyone who has had to use one of these can tell you it’s a complete nightmare and the whole Mac experience begins to fall apart.
Either they should have modified and shrank the OS for those models or shouldn’t have shipped such a bad model that only exists to milk people to buy the next one up
I have a non-mac keyboard, I had to unplug it before the latest update would get past the boot screen, that's never been an issue with the past updates, maybe that's what they experienced?
One of our mac mini CI server also failed to boot after update today. When it happens on Linux I usually try to troubleshoot the problem rather than resetting a whole machine. It does not seems to be feasible on macos with reasonable effort because of recently-introduced security features (and also my lack of understanding of macos boot process).
He skipped Catalina, that's a risky move. Updates have a higher chance of success if applied in the order they were tested by the QA teams of the developer.
That blog post is the biggest load of unadulterated bull excrement I have read in a very long time. It is pure unsubstantiated Apple bashing.
Get a grip man. An OS update fails on one machine and you find it perfectly reasonable to extrapolate this to Apple going to the dogs and that Apple management should resign ? What planet do you live on ? Your blog post is the very definition of overexaggeration.
Those of us who manage larger number of Apple devices have not seen any issues with OS X updates.
Sure there is always the possibility of an edge-case where an OS update can go wrong. But that can happen on ANY OS ... Windows, Linux, BSD.
Hardly surprising that people expect premium experiences after paying premium price for both software and hardware. That's why people are so easy to bash Apple, because they say that they are perfection. When something happens that are not perfection, people blame Apple.
You get the response based on the expectations you set. And with both price and messaging, Apple puts the expectations very high, while sometimes failing to reach them.
Because they do not include any kind of identifier for the user in the network data. Why would you store a big list of hashes? That would make no sense.
When something is so rigorously guaranteed that it's almost absolute, it's able to change your behavior, dramatically. It's incredibly liberating. It's like turning on the water tap, and never fearing that you're about to drink something toxic. You can do things quickly and casually, because you know you've got a solid bit of "terra firma" to stand on. If we didn't have things like this, we couldn't have i.e. airplane travel.
It's absolutely commendable for them being at such a high watermark for reliability, but - much like a doctor saving sick patients, or the mail service delivering letters, aspirationally speaking "nothing less than perfection is acceptable".
Glad I am not the only one who read through all that and was confused why they went to all the effort to write such a long piece about a failed software update.
One failure in 19 years is pretty damn good imho. I’ve had more macOS updates fail (and Windows and Linux and Solaris and OpenBSD) in that same time frame.
Sorry but this is a pointless post. I learned nothing other than the writer really seems to like to write.
As a counter point to his post. I’ve updated my 6 machines to Bug Sur and they have all been fine. In fact I would say Big Sur has been one of the “good ones” compared to recent years.
The point of OS X is that you pay a premium and now put up with a bunch of crap from Apple to not have to deal with stuff that "can happen on any OS." When you have to deal with it why bother? Why not run Linux?
For "pure unsubstantiated Apple bashing" that article is very modest though. macOS updates haven't failed for me so far, but I still notice a general decline in system software quality on macOS in the last 3..5 years, if the next OS update would fail for me, assuming that this is just another sign of a slow but steady decline would be very natural.
We've got a few thousand Macs and have seen a higher than normal proportion of kernel panics on Intel models since the very fast 11.15.x run.
So far, in addition to the usual panics on T2 devices from HID changes, the new ones are being triggered shortly after login to certain user profiles, especially when external video is connected.
No funny kexts, no custom drivers, just jamf and some very light profile and application level management. It's happening primarily to profiles migrated from pre-MDM / open directory bound Macs, but I only saw it in person yesterday when plugging in for a conference. Unfortunately, so did everyone else.
I think you are too “hard”, is a normal “surge” from a guy who has a trouble with a macOS update for the first time in 19 years! I also use macOS and I never had troubles with the updates but we can’t ignore that macOS in the lasts 6/7 years get updates a lot more frequently, so it’s normal normal have more troubles.
Surely Apple is not going to the dogs but is doing a lot more things/software/hardware and macOS is not “quite perfect” as it was 20 years ago. Also with the ARM chips Apple is doing a lot of confusion, especially with recovery mode/reinstall/format.
Linux upgrades don't fail that catastrofically. For example, a Linux upgrade can fail in a way where the GUI is not operable, but you log in into the console and fix it. Worst case scenario you log in in single user mode and fix it. Worse case scenario the kernel doesn't boot, you chroot from a live CD and you fix it. Everything is fixable.
The problem with macOS, or Windows, is that they are proprietary operating systems, if something goes wrong, the only way to fix it is to reinstall everything, that to me is not an acceptable fix.
If you're experiencing volume corruption, or can no longer guarantee the integrity of your system, at that point does it matter that it's open source? At what point do you cut the investigation, reinstall and get on with your day?
Proprietary OSes are less flexible but not to the extreme you're describing. In both instances there are faults you can recover from, and faults where you can't.
Edit, to rephrase my question as not rhetorical: I'm genuinely interested in knowing if troubleshooting this type of fault is a common occurrence to people outside of OS/kernel programming, as I lack experience with it.
The blog post states that the person has experienced a steady long-term degradation in the quality OS X delivers for them. I understand the failed upgrade was just the final straw before posting this writ.
And as a bystander, the post didn't even register as particularly harsh to me. It's what you write when you've seen things go down hill and you just hit another drop. Seems perfectly appropriate to me.
It has never failed here and once in 19 years is actually a big praise. BTW what's with those 15 minutes of preparing to update?
There are funny UX issues with Software Update. I press Restart now 3 times, nothing happens, after a while the system begins restarting and on the black screen I see a notification saying "We'll restart to update later tonight" or something. macOS is full of embarrassing stuff like this.
