Nasty macOS flaw is bricking MacBooks: Don't install this update
tomsguide.com
tomsguide.com
That, to me, is insane. It shows how shit the project management must be if they cannot even prioritize fixing critical bugs in the most important parts of the OS over releasing some useless feature that most users will never touch.
iOS keyboard input and text selection has been wonky for quite a while, and from what I've tried of Android, it doesn't feel much better on the other side of the fence either. Why can't Apple -or- Google get such basic interactivity right?
I also have a bug with their hardware keyboard. After a couple days of usage the keyboard can get unresponsive and the § key starts to magically appear in the next text field that gets focus, at least 10–20 times, as it would be stuck. Reconnecting the hardware keyboard usually fixes this issue, but it’s still annoying. The restarts that are required for software keyboard bugs are worse, though.
iOS is generally quite good at these things. Like any new platform, you can't expect it the UX to be identical to an old platform that you're more familiar with.
Is there a similar trick for selecting text? What about selecting text from another source like an email or website?
Next time: tap the cursor and hold. You’ll see it get bigger and you can drag it where you want.
ios 13 wants you to "grab" the text insertion cursor and drag it around -- but what are you supposed to do when the text cursor is not visible? and what are you supposed to do when an entire text region gets auto selected as in the url area ...
i would really love for someone to fix this issue
-- or if anyone knows how to do it ... here's the problem: say you have a long url that ends with a number -- and you want to change that number or select it to copy and paste it elsewhere. how do you do this on iphone in safari after ios 13?
I remember one of the main reasons I "jailbroke" my iPad was so I could install an improved/alternate software keyboard, but I remember reading they added the option officially a while back.
I'm curious: are they artificially limited or do they all just suck?
Every year, the phone OS has to keep up with new tablets and/or phones.
Every year, the laptop OS has to keep up with the phone OS.
That's basically it. At Apple, where the phone goes, everything else follows.
Fortunately I’ve got a 256GB SDXC card which I’ve installed Pop!_OS on - which has convinced me that now’s as good a time as any to extricate myself from the Apple ecosystem (which I’m all in on - iCloud, Apple Music etc). My next laptop will be non-Apple. First non-Apple hardware in this house since 2003.
I’ll try swapping the drive with my daughter’s - she has the same model. At least I’ll know whether it’s the drive or the logic board. Then I can sell the bits.
The usual build-to-order options like Dell or Lenovo being backordered due to the pandemic, I walked into a Costco and bought a LG Gram 17[0]. This is the first time I bought a laptop off the shelf, and the first time my primary laptop has not been a Mac, in over a decade. I installed Xubuntu 18.04 LTS on it with no fanfare and have been quite happy with it since.
The early signs of Apple's demise, indeed.
[0] https://www.lg.com/us/laptops/lg-17z90n-r.aac8u1-ultra-slim-...
I was kind of hoping it would be the logic board to self-justify a new laptop. Now I’m thinking I’ll get the cheapest 256 I can get and nurse the laptop along for as long as I can.
But I’m committed to switching to Linux. Still exploring distros.
I'm running a 6-year old MB Pro, updated within the first week of every releases' public beta availability, never reinstalled. I use sudo far too liberally to be good for system health, including (as long as it was possible) deleting all sorts of files in /System, /private, etc. when I run out of space and believe they're superfluous via cursory inspection ( i. e. sudo rm -rf /System/<asterisk>/{AMD,<asterisk>SCSI}<asterisk>).
And... it just works? I've never lost any data. I've typed in my WIFI password once and not changed anything since. Time Machine recently stopped working, and I had to follow the cable because I didn't remember where the HD actually was (taped beneath a cupboard, as it turns out; the power cable had somehow unplugged). Catalina did throw up a few permission dialogs after installation, but that took less than five minutes, even though I run a few low-level utilities (karabiner etc).
There are certainly some annoying UI bugs or inconsistencies. When you open an ePub in Books and immediately press cmd+f, for example, the search field loses focus repeatedly while the book opens. And why is cmd+f the shortcut for search almost everywhere, but not Keychain?
I recently installed Linux on a Desktop for some Cuda work, for the first in 15 years or so. Based on what I had read (mostly on HN), I expected Linux on the desktop to have made vast progress over that time. I'm sorry to say that it is, still, a complete disaster by almost any measure. I have no idea why the popular narrative sees the two anywhere close to each other.
Case in point: My installation of Ubuntu, using defaults, had an application menu with categories "Preferences", "Setting", "Utilities, and "System". Those all mean (almost) the same thing! There are two package managers now (apt and I forgot the new one's name), each with abpout five GUI and three CLI applications for installing packages. But both sometimes installed three year old versions of standard software (calibre, I believe). Good thing Homebrew is now available for Linux, because somehow they have no trouble keeping up.
Font sizes, margins, and contrast are all over the place. There are dozens of color schemes, yes. But much like unhappy families, each one fails in its own manner: with both a dark and a light one, I encountered parts of the UI where you'd suddenly have light buttons with light (invisible) text. Or, IIRC, light text on dark backgrounds. Which works -- until there's a link in the text.
There seem to be five different subsystems between my mouse and the desktop. None of seemed to allow adjusting scroll wheel acceleration. (Even pointer acceleration felt wrong to me. And even though I found the system of partial differentials that seems to be the official interface for adjustments, I failed to find the sweet spot. But this might be an issue of familiarity more than objective quality).
Good luck with that.
Apple used to maintain a 2 year (or longer) cadence earlier.
I suspect the first few months after each release used to be focused on hardening it. With the yearly release cycle they probably have to shift over to working on the next release almost immediately.
If the same thing happens today where everything is soldered and epoxied together, you practically lose the whole machine and all the data on it. Woof.
This article makes me reevaluate that thought.
Coincidentally, one of my colleagues had warned the team about installing macOS updates, since he was having repeated issues with kernel panics on the original 10.15.4 update.
I hope Apple can get some of this together a little better going forward.
However, I'm still having kernel panics after the Supplemental update. But interestingly, they only happen when the machine is supposedly sleeping, with the shell closed. I kid you not. I will have it sitting there on my end table in the living room, shut for over an hour, and all of the sudden it will reboot with the lid closed (I know because I have the startup chime enabled). I've never had a Mac kernel panic while supposedly sleeping. This is so f*cked.
I appreciate the transparency here. Outside of API documentation, a lot of plain-language information doesn't go in-depth enough about MDM capabilities.
http://launched.zerowidth.com/ makes it easier.
In this case, the kernel panic seems to be related to the Intel Iris video driver.
[1] Which is still terrible, these things should just not happen. Bad quality control from Apple' side and inexcusable for > 1000 Euro 'premium' computers.
Major version upgrades are an entirely different matter, although even they sometimes just work with Clover.
Is there some reason why phased release isn't more common among OS/firmware updates?
Auto update does not have to mean "update my device the day new versions are out".
Edit: Is this related to public disclosure of security bugs? If so, the community should change their standard so that public disclosure doesn't happen until a week after the update is available. This would allow for phased rollouts.
The update being available is public disclosure! For an experienced reverse engineer, comparing the files before and after the update is usually enough to pinpoint the security issue.
If there is an obvious to find, easily-exploitable bug, the update could be pushed to everyone (the way updates are pushed today).
There is a balance to strike here. If issues with auto update cause enough people turn auto update off, I'm not sure it's helping security.
Finally - Apple already has a beta programs including public betas. Are those binaries not already used to find exploits?
Essentially, I'm not convinced a phased update (with public security disclosure slightly later) would lead to worse security than we have today.
Releasing a new major version of your OS yearly is insane.
On macOS you can stay one major release behind. So, stay on mojave until the successor to catalina is released, then upgrade to catalina. You still get security fixes and full software ecosystem support, but far fewer issues with buggy updates.
On android and iOS you don't really have good options to my knowledge.
Wikipedia is indicating two major releases have support, not just one ("End of support date" in the table):
https://en.wikipedia.org/wiki/MacOS#Release_history
As a data point when doing stuff in a High Sierra macOS VM earlier today, Homebrew still seems to be making binary packages for High Sierra too.
So the "two releases behind" seems right.
I feel pretty sure this must be common because I still see apps and infrastructure, even last month, coming out with updates related to Catalina compatibility.
It sucks though. I have a Google Pixel phone and never hesitate to apply updates, I even look forward to them. I don't think I've ever had an Android phone experience the sort of severe regressions that occur regularly with Apple updates.
I've been a MBP user for a long time, but I'm really starting to wonder if it's time to take a leap soon. I've been using this laptop for maybe a year and:
1. I'm scared to upgrade to the latest OS
2. The screen's coating has been damaged and has key-print impressions on it. This always happens and Apple seem incapable of fixing it.
3. The latest butterfly keyboard has at least not broken on me for 8 months or so, but the options, C and A keys have worn through and now have holes in them. This started happening a few generations ago and now seems like a regular problem that I'm just supposed to accept.
4. Very soon after getting it something happened to the metal such that it has a large discolouration on the bottom left, of a type I've never seen before. Nothing seems to fix it.
5. The screen routinely gets a yellow splotch in one of the corners if the laptop has been in my backpack for a while.
The problem is that despite not even really having improved for years, macOS is probably still the best OS out there. Linux on laptops has never worked well - my colleagues who try to use it routinely have issues with webcams not working properly. Windows laptops seem to vary wildly in quality and many of them have stupid design flaws like putting the webcam at the bottom of the screen instead of the top, coming loaded with crapware, anti-virus products that cripple performance, Windows is still a rats nest of weird problems under the hood. And their way to make it better for developers is to just bundle Linux?
I really wish there was more competition in the laptops-for-technical-people space.
If you want good linux support on a laptop you need to do some research, but there are options. Thinkpads, XPS, ... They’re just not that much cheaper than a macbook if you want comparable specs.
Been working fine for me since October 2018 for dual boot gaming & firmware updates. It will update to the latest version of LTSC when you start up.
Works with the latest hardware too, I am using it with a 5700xt (released last July, purchased in December) and a Ryzen 7 (also released last year). Only thing missing is the new Microsoft Store, if you search for it there's a repo on github which will let you install it.
For all the Windows 10 ISOs you need to use Rufus instead of normal dd/cat the ISO to a USB stick (doesn't work with dd/cat). Rufus works in a VirtualBox VM with USB passthrough. During the setup, select that you don't have a license key and then activate it later as mentioned in that post.
Very happy that I've helped someone.
Is Microsoft concerned users would like it too much? Maybe that’s a sign that their constant feature updates aren’t actually desirable and they should back off.
Maybe they shouldn't have released 10.15.4 right now, but instead only the security fixes in this release. But they had to push out that release to support the new MacBook Air.
Another problem with such reports is that it is hard to get an idea of how widespread this problem is. And whether the problem (perhaps) already occurred with previous Catalina updates, but is now picked up more.
1. As seen here, it's a bugfest.
2. For user-facing features, it brings…Sidecar? And a bunch of half-assed Catalyst apps, I suppose.
3. Such an exciting new set of security features—really met what I have to guess was their goal to quickly train me to click OK on every single dialog to get to work.
I'm a consultant and talk to multiple clients all with their own video conference solution and having to do this every time I join a call where a client is using a different video conference solution is very unprofessional. The worst was using Teams where I had to close my browser before I could join the conference, after I had already queued up everything I needed for the call. So now everyone is waiting on me to re-open everything and reset the state for the demo.
I never thought I'd miss Windows UAC.
I assume every update ever has messed up a handful of people's computers, even if just coincidentally (e.g. the computer was going to die on the next reboot no matter what).
If thousands of people were complaining on Twitter that would be one thing. But since there aren't, and considering that Apple hasn't pulled the update...
Is this just much ado about nothing?
I think it's fair to say this issue is relatively rare, but there are definitely more than three reports of it happening. Heck, some people in this thread chimed in to say they got bit the update.
It seems silly if you're not affected. But now that it's hit me, i'm about to go back to mojave. If it was just a weird bug, I could deal with it. But to kernel panic is a big deal to me. I cant trust the computer to not break if I walk away from it for 10 minutes.
I guess I could turn off sleep, but that doesnt seem like a solution to me. To keep it running all the time so it doesnt break? nah.
I updated two machines myself with no problem. Guess I am just one of the lucky ones.
It was extremely satisfying to finally isolate it to 100% reproducibility. Peak of my career right there.
"This supplemental update for macOS 10.15.4 Catalina, released Wednesday (April 8), was meant to resolve issues created by the 10.15.4 update in late March."
¯\_(ツ)_/¯
And by “bricked” I mean - no signs of life whatsoever
On a serious note, possibly because things are much too capable and complex than they used to be, combined with tight release cycles.
1. The OS update policies became more "agile". Both Windows and macOS starts to update the OS more often than ever. macOS moved to an annual update from biannual, Windows 10 switched to a subscription-like model.
2. They laid off the testing team and stopped testing on actual PCs in favor of VMs. [1] It makes sense considering the broken updates are often more low-level bugs that cannot be caught by VMs.
[1]: https://www.techradar.com/news/windows-10-problems-are-ruini...
Must be good for the bottom line.
The cost is absorbed by the unlucky users.
was meant to resolve issues created by the 10.15.4 update in late March
The "fix some bugs, add some new ones" trend has certainly gotten more common within the past few years. I still remember when the official advice was not to install updates unless you were experiencing the issues it was to fix, but that wasn't really applicable to security fixes, and now official advice seems to be "always install" regardless of security or indeed usefulness.
Isnt that because lof of companies are eliminating QA people and expect the devs to do the testing.
your probably not missing much, there are some serious regressings in that release, for example [0]
[0] https://forums.swift.org/t/swift-5-2-struct-property-wrapper...
(for me, cant make release builds because the compiler crashes... sigh...)
https://forums.developer.apple.com/thread/130988
So now I’m stuck on 10.14.6 and iOS 13.3.1, one of which is not getting security updates.
Really hoping Fuschia or someone steps up and displaces Apple.
I can now debug 13.4 devices on Xcode 11.3.
They don't need to ship anything new and they have the perfect excuse to focus on improving both their OS and their processes that allowed such flawed things to reach their users.
Usually fixing bugs requires access to specific repro machines, and bricking might require access to a lab with jtag style hardware.
makes me wonder what the realities of telecommuting really are.
Right now it seems the only way to do that is wait a long time before installing any update they push.
You absolutely can do regular updates that are high quality. You just can't do them in an environment that's deadline driven, or which gives PMs too much power over what dev teams work on.
But you do have a point: it should be product driven not feature driven.
https://support.apple.com/en-us/HT210650
Although this change may not be immediately visible it does increase security a fair bit (malware cannot embed themselves or replace system files) and also prevents accidental deletion.
Apple made large strides in the last few years improving security, e.g. by introducing SIP, but also by offering fine-grained permissions for camera, mic, etc. access. APFS created its own road bumps, but was a long overdue replacement of HFS+.
IMO there were also some nice user-visible changes, such as dark mode, replacing iTunes by several separate applications, and dynamic desktop.
I agree that there are also superfluous changes or steps back, I am still grumpy that they replaced Spaces by Mission Control. I don't care for all the deep iCloud Drive integration and think it is unfair to the competition.
I strongly disagree that there are no useful new features the last five years. The problem is more the lack of quality control. That said, I started using macOS when 10.5 came out, which also had its fair share of ugly bugs (e.g. serious problems with maintaining WiFi connections).
I agree some of those security features are good to have but it's like they forgot about adding anything worthy from users' perspective.
This was nice, though it has to be said that the disk space reduction was not because of real optimizations. They removed support for PowerPC in 10.6, so the included applications are not fat/universal binaries anymore. They also removed printer drivers and Rosetta (binary PPC emulation) from the default install (which became an add-on).
Of course, still nice nonetheless if you didn't have a PPC Mac or didn't use PPC applications.
For 3rd party apps universal binaries made complete sense as a way of simplifying app distribution, but that doesn’t extend to the OS.
Heck, do Intel Windows 10 installs include ARM copies of every binary now? I hope not...
Can you elaborate? This terrifies me as I rely a lot on Spaces to organise my desktop between different projects
Is this a change in Catalina?
Is seems like Spaces is still a thing: https://support.apple.com/en-gb/guide/mac-help/mh35798/mac https://support.apple.com/en-gb/guide/mac-help/mh14112/10.15...
I'm pretty happy with how this works in Mojava, apart from the major gripe that the apps don't restore to the Space they were in after you reboot.
No this was done years ago in 10.7. Spaces was awesome and super configurable. But they combined it into Mission control and made it much simple but less useful.
That's very much despite Apple.
I did a factory reset via finder and that worked. As soon as I restored the iCloud backup: back to the white screen.
I had to revert to 13.4 and could then restore the backup. Now I’m afraid to update again … I guess I’ll have to wait until 13.4.2 or 14 are released. Luckily 13.4.1 does not seem to contain security relevant fixes.
I'm still getting freezes on my brand new 2019 MBP due to the still unfixed GPU switching bug and I don't dare install any macOS updates now.
Apple has had a lot of software quality issues in recent years but this is a new low.
Initially, I tried to restore the system via Virtualbox where I installed Mojave via Apple Configurator but that didn't work as it break during the revival of the last step because of the constant USB de-plugging from host/guest machine during the process. I think just the revival step in 2.12 should also work as it flashes also the bridge os and power it up which should avoid any data loss but don't take this for granted.
Installed Snap Camera a few days ago, was on a video call and when I came off it the green camera light stayed on. Even after quitting Hangouts and Snap Camera.
I restarted and the green light persisted throughout the whole reboot. I then shutdown, and it finally turned off.
After turning it back on, the camera didn't work at all. I uninstalled Snap Camera, reset PRAM, SMC. Nothing seems to fix it. Now thinking it must be a hardware fault, but think I was on the latest update already and just seems like a strange coincidence that it happened almost immediately after I used Snap Camera for the first time.
I've done the same update on my Mac Mini 2018 and now the fan has stopped working after reboot. Tried to reset NVRAM and SMC without any luck... the fan is still unresponsive. Apple diagnostics even confirms this by giving an error code PPF003
It says it's installing, I reboot, it does some stuff, then I come back and the update isn't applied.
Quality stuff!
I'm aware of security patches etc, but risk of regular things breaking for little apparent upside has always led me (and most folks I know) to avoid upgrades except in the above-mentioned case.
Edit, not just asking about Windows, but all operating systems, OSX, iOS, Android, Linux, etc.
Windows 10 seems to be pretty good at keeping its users on the latest version? (Unless you actively tell it, not to?)
So a driver bug, affecting a single external device, is worse than completely bricking an entire machine?
Microsoft have had their share of controversies with updates (recent memory is deleting user data folder before they halted the rollout), but they have never made a machine unresponsive to the degree of the multiple anecdotes in this comment section.
HN has some smart folks, but the leeway those folks afford Apple despite their continuous hostile and incompetent behavior is incredible.
Having said that, I may have some kind of mental disorder because the thought of being behind when new software is released is abhorrent to me... Oh, but I don't AUTO update. I like to do it manually so I can read changelogs when available.
... I'm starting to feel like that person who enthusiastically tells everybody how much they love to peel their dead skin off. And everybody gives them 'that look'.
I won't say this is not an issue specific to this supplemental update, but I'm not buying it is _yet_.
It looks like one of the usual consequences of a system update going wrong. See that the user reports that after fixing this through Recovery Mode, they were asked to install the very same update again, as if nothing had happened. This looks like a rollback.
I wish MacOS was more transparent about what actually happens on those cases. If you had to roll back, let me know. If you could not install because of any reason, let me know.
Steam almost always already knows if the game has a '64-bit macOS depot' configured -- ie ships 64-bit binary for macOS -- though in some cases the 'depot' may be configured as "macOS - All" in which case Steam may not know for sure.
When Catalina released, they defaulted the checkbox to "No" regardless of the game depot configuration. Any remotely recent game that supports macOS almost certainly is shipping a 64-bit binary (been quite a while since OSX was 32-bit only) -- so odds are very good that games supporting macOS will work with macOS Catalina, at least likely for anything in the last 5 years if not longer.
In our case, it took several weeks to find the checkbox to update our store page. In all fairness, Valve sent out an email alerting of the concern ahead of time; we just didn't read it thoroughly enough to realize that the default setting would be "No" even when explicitly having a 64-bit macOS binary configured.
Apple killed them all for no reason.
You hear these complaints with every macOS release, but they are probably a very small vocal minority. I've never run into any of the reported problems nor have my 20-odd colleagues.
I’ve been avoiding Catalina because of report after report of how buggy it is. For example, emails disappearing and can’t SSH on a custom port are still problems even with 10.15.4.
OTOH, my work machine (a 2016 Macbook Pro) is stuck on Mojave (due to incompatible third-party software) and the latest "Security Update 2020-002" broke video conferencing. Right in the middle of a pandemic forcing everybody to work from home!
Of course, this being Apple, I'm not holding my breath for a fix anytime soon. They don't seem to release individual updates anymore, and even these bundled "Supplemental Updates" are few and far between.
Apple really needs to get its act together regarding macOS. People need macOS to actually be productive, unless they'd rather lose this segment completely to Microsoft.
Still bad that it's happening, but the term "bricked" is thrown around so flippantly these days. If a piece of tech is bricked, it's not recoverable at all.
It just changes the noun brick into a verb. So if a laptop has bricked, it means it’s like a brick now (useless)
I kinda like "cold bricked" personally.
- soft: need some sysadmin mojo, USB adb etc.
- hard: can be fixed with soldering, physical repair of sorts.
- cold: will never again work even if you're Louis Rossman (maybe as a ship of theseus). Think CPU fried.
https://www.techopedia.com/definition/18959/doorstop
But you can't keep a door opened with a broken cell phone, so they had to invent a new mobile-friendly term "bricked".
For a mainframe or minicomputer, you could say "monolithed" or "bouldered".
Anyway, the "bricked" expression is more commonly used in smartphones and PlayStation Portable (PSP) than laptops. Then it makes more sense when considering the size of the devices.
I think it's also at least in part more commonly applied to those types of devices because they're generally more locked down than a laptop, leading users to risk bricking by trying jailbreaking them. Installing a custom OS on most laptops just isn't nearly as risky in terms of bricking potential.
After that the phones would be, essentially, a brick.
I did not have the technology to fix it - a JTAG connector and software to drive a reinstallation of what I assumed was the firmware.
More generally, any issue with a computer can be fixed by a sufficiently skilled and determined user - maybe with some data loss. Even a broken firmware might be fixed by writing a new one, cracking the authentication mechanism, and building a custom rig to flash it to the chip.
It's all a matter of degrees and how much users can be expected to do which makes defining exactly what "bricked" means very difficult.
It requires a separate, recent Mac, involves arcane combinations of specific ports, keys, and timings, relies on luck and voodoo, and ultimately may not even work.
For some consumers, if recovery mode doesn't work that might be the end of line.
However, there are more advanced users that can access the storage via another system and fix it there, hook up via serial port or JTAG and flash new firmware, or even desolder and replace chips.
On the far other end of the spectrum, there's going to be people that consider "doesn't show the normal UI after I press the power button" as bricked.
I'd like to keep it this way even to the general lay person. Because if you need hardware the barrier to entry goes up significantly.
Agreed that 'bricked' is not the right term here.
From Dictionary.app that comes standard with the Mac: "a smartphone or other electronic device that has completely ceased to function"
But because they're "captured usage", they're generally what people would understand by the term.
"Literally" is often used when the subject is metaphorical, but that's not the same thing as it meaning "metaphorically". If you took the "literally" away, it would still be understood to be metaphorical - the "literally" is intended to strengthen. It's the normal sense of literally, used hyperbolically.
In the same way, when someone says "you left me waiting for days" we don't say that "days" sometimes means tens of minutes.
The Sonos "recycle mode" bricks the device. If same effect is produced by a bad capacitor (rather than a deliberately-blown efuse), it's just broken.
>For some consumers, if recovery mode doesn't work that might be the end of line.
If you want to go by that rule, for a large number of people simply getting a virus would mean their computer is bricked. I think it's safe to assume "bricked" means it's a hardware level issue.
Failing to make a back up of the software on a device which you're attempting to hack and subsequently destroying hardly seems relevant to an auto update locking you out. That's more like changing your password and forgetting it, or back in the day formatting your hard drive without a copy of XP to re-install. I don't know which phone you're referring to, but every phone I've rooted has been salvageable.
Especially since countries other than mine have called out a complete lockdown. You can't just walk into an Apple store right now or call a techie friend over.
Chrome and Firefox, as well as Microsoft, have announced that they will only distribute stability fixes to prevent people working from home from having trouble.
In any normal year, I'd say "shit happens" (or "that's the risk of buying Apple", given that they don't even offer to do pickup of your >€2000 machine like most businesses selling laptops in that price range do). Right now, any disruption in computer performance is debilitating.
Apple needs to confirm these cases ASAP and/or pull the update for the devices affected. Apple isn't the only one releasing bricking updates by the way, Samsung has pulled an update for their A70 phone because of bricking issues two days ago.
In this case, you can probably recover through the DFU mode and reinstalling Bridge OS. But this requires another Mac, which some people may not have, and at least some tech know-how.
According to postmodernism, everything is like that.
Maybe, but this sounds really strange to my ears. I think for me the difference between bricked and non bricked is the expense of the repair, and maybe if it's more than the value of the device. If the wiring in your house is all melted after lightning strike then it's bricked. If you total your car it's bricked.
And yes, there's a difference between a soft brick and a hard brick, and it depends on your ability to repair.
Getting my phone stuck in a bootloop is usually a soft brick, but my Pixel went into one a few years ago for no apparent reason (likely mobo failure) and that was beyond my capacity to repair or diagnose, so it's bricked.
Hardware errors are hard bricks.
This isn't a difficult delineation.
So flashing new firmware to fix the problem, not bricked.
Requiring soldering stuff and replacing hardware, it's bricked.
Otherwise, as you said, 'bricked' can mean almost anything. So what's the point of using a word without any particular meaning?
Not all people own more than one system.
If it was my only computer, and depend on it to do work and earn my living, and it broke, I would be terribly upset. Especially these days with the quarantine.
It makes a mockery of the idea of advice. Imagine your parents calling you and telling you that their machine is bricked. Well, I guess they must buy a new one and the old must go to electronic recycling. No need for a diagnostic of any kind, it's bricked.
Linguistic descriptivism will be the death of communication and standards. Only pushing back at the creep will allow people to discuss anything technical. Imagine the trend should it overtake medicine. If a certain kind of stroke is suspected, the patient should take aspirin right away. No, aspirin-aspirin, not ibuprofen. Then you have someone who says, "Well, ibuprofen is an aspirin sure!"
Even within a field, people may have slightly different interpretations of what a word means. Take consumer electronics. Say someone plugged in the wrong power supply and blew an inductor. If you don't know what a power supply filter is, it's garbage. If you do know what's going on, it's a simple repair. Now look at the case of a bad firmware update, where the device cannot reach recovery mode. Someone with a knowledge of embedded software development may know how to fix that, particularly if they have connections with the vendor. Most people will consider it damaged beyond repair. The thing is, you can have one person who is knowledgeable about software and another about electronics having entirely different interpretations of what bricked means. The only real commonality in the definition is, "I cannot repair it." For someone with negligible knowledge about how consumer electronics, that is going to have a very broad application.
Personally, I think that bricked means that it can only be repaired with specialized tools (including soldering irons or JTAG interfaces). But hey, I'm sure a lot of people would beg to differ.
EDIT: missing don't.
Machinery standards at least define this mechanically as a screwdriver or above. The software equivalent may be a USB cable.
(FWIW: If the terms are to be interpreted relative to the device owner's competence level, we also need to prosecute people saying "You're dead to me" for murder)
“I bricked my phone and had to spend a whole day reading obscure guides to three different levels of embedded software before I could get it working again” feels like a perfectly valid sentence. You can fix a bricked thing, just like you can fix a broken thing, but both require some investment of time and possibly the use of tools that you won’t usually find outside the hands of specialists.
imo "Bricked" means you turned your piece of electronics into a brick, it cannot be recovered. Otherwise as mentioned before, it's just "broken", which can be recovered from in most cases.
A bricked device usually can be recovered using only software tools and sometimes a special cable. Sometimes those tools or cables are only possessed by the manufacturer, which frustrates consumers and makes it seem like the devices are unrecoverable, but they’re not.
A broken device, on the other hand, can’t be recovered using software tools or a special cable because it contains broken parts that must be repaired or replaced.
A bricked device has a slight chance of recovery, if you have the tools/skills/training. It is a brick until that repair with high applitude is completed. Something that is broke doesn't mean it is functional or not, just not at a perfectly working condition to worse. it might be repaired by doing a reset of a device, or something more advanced.
This is from my experience dealing with non-technical people who are mechanically inclined, but not technically inclined. they will call with a "broke" device that just needs a reset since they have too many users in a system all attempting to run the same device on different things. (Sorry keeping vague to keep me out of hot water). They will also call in with something "bricked" because the device won't function due to a damaged USB port, and they don't have the skill set/components to solder a port on something electronic. And then further down the scale it is a paperweight when it won't boot and is a piece of hardware they hate.
The threads here show that even highly technical people disagree on what conditions should be considered bricked versus broken. To a non-technical person whose device isn't working, however, there is no practical difference.
I would refine my definition to be that a "bricked" device is something that has occurred via a failed software update making the device inoperable to all but the tiniest subset of users.
It’s fair to say “bricked” is a sub-category of “broke” but if something is software repairable then it isn’t bricked.
Also aspirin and ibuprofen are very different chemicals. I wouldn’t advise conflating the two. The resulting effect could be worse than the brick/broke problem we are discussing ;)
The term unbricking also exists, which directly contradicts the idea that a bricked device cannot be recovered.
For example - years ago, NVIDIA got in some hot water over some of their mobile GPUs getting so toasty some of the solder points broke, leading to bricked systems. If you were the adventurous sort, you could heat the laptop enough to reflow the solder and end up with an unbricked system, but that's not an approach most people are going to take.
However, I think the situation you might find here is that, if you define "bricked" as a situation which is literally unrepairable, almost nothing is ever bricked. Which similarly makes it a not very useful term. Bricking would be a term exclusive to devices which have been...exposed to an intense EMP, or run through an oven. Situations which, while they may occasionally happen, happen so rarely that no one would ever use the word.
You can always bring it to a repair-service.
I'd say it is "bricked" if repairing it costs more than buying a new one; i.e. it's a total loss situation.
There were some "unbricking" guides but those usually required hardware voodoo, soldering, etc., things that most people couldn't do.
Now it's a clickbait term that has lost its original meaning of "you now own an expensive paper weight".
Seems pretty standard (and it's how I use the term[s]).
> No power at all. I tried SMC and PRAM reset but sadly was unable to do either since my computer won’t even turn on.
[0]: https://forums.macrumors.com/threads/warning-macos-catalina-...
[1]: https://forums.macrumors.com/threads/warning-macos-catalina-...
If so, maybe a DFU restore is needed.
I wouldn't be surprised if Apple ships Configurator iOS someday, but today is not that day.
Still sucks, just a bit less. not really different than an iPhone, except that you can rebuild those on Windows (for now).
The checkra1n team recently showed that all Macs with a T2 chip are vulnerable to checkm8 exploit used to jailbreak iPhones, and this could persist for a while due to the T2 chip staying on between application processor reboots.
This allows override of mic disconnect (except on the newest models which switched to hardware disconnect), Secure Boot/Firmware Password, and allows you to bypass Apple's signing of Intel ME firmware, TB3 firmware, and CPU microcode. Whether or not there is still Intel signing after that is unknown, but there are already some sort of issues with the host key being leaked on that. The one useful feature I can think of is allowing SSD replacement (you still have to find a way to resolder ofc) and Touch Bar customization.
I think the most likely attack vector for this is an evil maid style attack where corrupted T2 firmware is loaded that rewrites the contents of the SSD while macOS is running to launch further exploit code on that platform.
The one thing not to worry about is if you have FileVault turned on with a password, that still can't be cracked because your password is not stored anywhere on the device. But the BitLocker-style automatic encryption with no password that just locks the SSD to that specific T2 chip isn't useful anymore.
Took a day, but all's well now.
Glad it helped.
No: There’s no opportunity to use FileVault recovery keys when you’re doing a DFU on a T2 Mac. If you have to DFU, your data is lost.
(edited to clarify per reply)
It’s not failure, it’s secure by design to prevent attackers/governments from stealing your files without consent.
Under DFU brick and reset circumstances, the private key is gone, because otherwise an attacker could just upload a hacked firmware via DFU and access all your files.
I assume the installer uses a different process that performs a DFU upgrade-in-place that safely manages the handoff using signed code and such, but that’s not the process we get as a last restore described above.
If you don’t have off-device backups, you’re accepting the risk of losing all your data at any time due to any number of possible failures (software and hardware). Not much use getting upset about this specific case.
Computer went to sleep after the update. This was on Friday. Haven't been able to do any kind of thing on it. Trackpad 'clicks' but no screen, not touchbar, nothing (2018 Macbook Pro).
Called apple, they tld me it had to go a store, unfortunately where I live, due to quarantine there are no shops available to pick it, and we aren't allowed to drive outside our district.
Will try the apple configurator idea someone posted below and will report back. Just need to find a cable that supports both power and data that is USB a to USB c.
Tomorrow will call a close enough store that seems to be (stated on their website) doing pick-ups and see if they can come and pick it.
My main concern now is they will try to try and charge 1000 or so euros for a repair and I will have to eat the cost as I can't really ask for a replacement one from company or friends since everyone is stuck at home
Probably a bigger issue isn’t that they’re a single laptop failure away from collapse, the bigger problem is that replacing a laptop right now would be unusually difficult.
“A small number of machines” is not what we are talking about. We’re talking about one single consumer-grade laptop that would collapse the entire business forever.
Now, you really can’t buy another laptop same day.
Bricking does not mean unrecoverable. That’s why the term unbricking exists.
It’s absolutely ok to call something that most regular users will be unable to fix on their own and in the meanwhile their device won’t even turn on “bricking”.
I do however note that when adjusting volume in the status bar the UI dragbar position is very jumpy, which gives a rather unpolished impression.
I've noticed this as well. Qualitatively, it seems to be worse when external audio output devices are selected, than for the Internal Speakers.
And it went on to happen in the loop: login -> white screen -> reboot.
Recovery mode + macOS reinstall = "Can't install macOS on this computer"
To my that starts to match the definition of "bricked" already, because a normal person really can't probably recover from this at that point.
Recovery mode (booted via USB drive) = "Installer is missing assets" (tried 2 USB drives, after re-downloading Catalina installer
Recovery mode (booted via USB drive with Mojave) = this is not supported anymore
Recovery mode -> Disk Utility -> First Aid -> run a FS check, turned out APFS volume is corrupted ("fsroot tree is invalid")...
Target mode (went to BestBuy to get a special "SS+" certified USB-C cable, because of course a normal cable doesn't work) connected to my macOS Catalina Mac mini -> couldn't unencrypted the APFS partition enough though the password was correct. Read every possible target mode tutorial, and there's no mention about what Catalina did to Recovery Keys/User (other than them simply not being there), but you need to pass a UUID of a recovery used, and I only had some iCloud Recovery Keys there. Tried my MacBook password, my iCloud password etc. Nothing worked. Basically I couldn't un-encrypt the drive.
Ended up just starting Recovery Mode again, mounting a separate USB disk from the Terminal, copied valuable stuff over.
Removed all APFS volumes, tried to re-install on a fresh volumes. Nothing worked. Was getting the same issues from the installer. Gave up. Packed this $4k computer and sent it back to Apple. Writing this from my Mac mini, with Catalina, happily rejecting any updates. BTW, Catalina installer is a joke. It is broken since Catalina came out, and it's not getting any better.
But you're still absolutely right. They should be testing every possible configuration.
Tomorrow I do a hard reset. Nice way to spend my Easter holidays.
I still prefer them for now, but eventually the cost/benefit ratio just won't make sense anymore. If I'm paying for crappy quality, I'd rather just buy cheap.
I've been protected from a major flaw by another, equally major flaw.
I guess it's better than the alternative of getting hit by them both.
https://mrmacintosh.com/10-15-4-supplemental-update-bricking...
It seems Apple is finally getting their shit together on the Mac hardware front. Now they need to do the same for macOS.
Moving to yearly macOS releases was a bad idea.
The two events are likely unrelated but so far has not failed me.
Mine is working so far
Sounds like a real pickle for those affected, though. v_v
Linux or Windows, doesn't matter - you can always recover via BIOS.
How can Apple users live with crap like this?
This issue is exactly why I only occasionally install updates, and if I do I wait a while for the canaries to leave the coal mine. I've been a user of macOS, distros of Linux, and Windows for years, and in every one of them there was the inevitable update that made the OS unusable. It happens rarely but, when it does, it's a miserable experience.
Otherwise Linux is pretty damn rock solid. These days "it just works" applies to Linux more than any other OS.
And with the new Nvidia-less Dell Precision 5400 the system runs cool and quiet to boot.
Linux isn't what it used to be...
Granted it wasn't strictly "bricked", but I think no-wifi meets some of the bars for unusable.
So yes, bad updates do happen.
https://linuxreviews.org/Linux_Kernel_5.5_Will_Not_Fix_The_F...
They should be fixed by now, but basically the latest LTS kernel was broken for 3-4 months on many Intel GPUs.
On a PC pretty much the only thing that can brick the machine is a BIOS update.
This auto-updating treadmill we've been sold on is really beneficial for developers, but it causes some major problems for end-users.
I have honestly lost track of the number of times I have had an update cripple a device. My iPad got ruined, my PS3 lost Linux, iTunes removed p2p music sharing in a deceptive update, Cisco gutted residential routers and forced people onto their crummy new platform[2]... I could go on and on. And this isn't even counting Android. I had a friend lose Miracast support in an Android update, minutes before he had to give a presentation.
And those are all examples of updates functioning as intended, even though they are against the interests of the users. I can add more citations if anyone is curious.
So no, if you use your device for anything important, then don't update it until you've made a backup, and had a chance to read up on any potential problems. It really says something that in this day and age, I've had to become much more vigilant about backing up my data than ever before.
[1]: https://news.ycombinator.com/item?id=21064663
[2]: https://tech.slashdot.org/story/12/06/29/1425210/cisco-pushi...
How so? If anything, I think it's just as bad for developers, who expect a stable foundation to be able to build on.
One thing I really don't like about it, is how it forces you to spend more money on the latest hardware you don't need. Something is broken when I can't use a three year old phone to pay a parking ticket in my city. And they can just scoff at me and say that it is my fault for daring to go two years without buying a new device.
Being evergreen means that any vulnerabilities in DRM and platform integrity are swept away under a tide of updates. As long as you are constantly pushing updates, and making it very hard for people to use an out of date version, you don't have to worry nearly as much.
Finally, in the bigger picture, it strategically makes sense because it keeps your competitors aiming at a moving target.
Why do you think it is easier to install an older operating system than it is to install an older version of Chrome? Do you really think it is solely about your security and safety? Of course not. They don't care about end user security or privacy nearly as much as they do about keeping their platform secured. In fact, it is still easy to buy out a trusted Chrome extension and attack users. There is still no possible way to disable extension updates. Google knows about this and they won't change their architecture in any major way; they just keep the updates and revocations rolling and rolling, and any users who are hurt by this are an acceptable loss.
I feel like Microsoft somewhat had that problem solved --- if I just want to create a basic Windows GUI app, and know that I don't need any of the newer features, I can create one binary which will work from Win95 onwards.
On the other hand your developing environment can suddenly change, what worked fine yesterday requires a complete rewrite the next day.
I've seen that happen with most of the other platforms; and relevantly to this article, Apple is a good example of constant breaking change.
The Apple updater system can target specific machines (not sure if it’s by serial or MAC address or what).
I do not want point and click military intelligence RCE capability on my machines.
1) they take ages to complete and I want to run them on my schedule instead of being forced to take a 40-60 minute chunk out of my schedule. 2) there's always a risk of an update breaking something and if that happens, I'm OK with being a few days/weeks behind.
In recent years, big OSX updates are effectively a beta release until the second service release or so. Also, they have been enormously underwhelming in the sense that I can barely tell them apart. There's generally very little in them that I need/want. I don't use any of the bundled software (mail, icloud, the office stuff, etc). All I care about is being able to run my browser (Firefox), Homebrew, some games (quite a few less as of Catalina), and development tools like vs code, intellij, docker.
I actually updated my imac to the latest Catalina last weekend, and applied the fix we're discussing here a few days ago to be able to run the latest x-plane beta with a new rendering engine that uses metal (and vulkan elsewhere); it complained it needed the latest driver. Didn't break anything for me luckily. My mac book pro is not getting the supplemental upgrade until I know it's not going to ruin my day.
Just disabled auto-update. Phew.
Edit: there's a Configuration dir in Recovered Items which has two other dirs, a private dir which contains another etc dir which contains a shells default file. Nothing else.
Unsure why this didn't help me at all.
Bravo.
As a result, some large percentage of readers are let down by the actual content, and think it wasn't worth reading, let alone commenting on.
That your Mac can get to the point by an Apple OS update that you need to enter recovery mode? (Even as someone who is exclusively Apple, I have seen the mockery Microsoft has endured for similar).
Or worse, to the point that unless you have access to a spare Mac or the _currently closed_ Apple stores, you are shit out of luck?
"Eh, no big deal, not worth reading or commenting on".
The fact that an OS update can brick a device is a non-event, it's happened before and will happen again. Apple is not perfect, no one is. They do a pretty good job most of the time, comparatively¹.
The fact that in the case of Apple you have to suffer the aforementioned shenanigans to maybe solve the problem is, well, coincidental. It's not an event in itself either.
Note this is why production machines in professional environments usually wait a few weeks-months to update. Also why security updates are generally offered on their own (shouldn't wait to install those).
____
1: Note that, I personally can't stand being at the mercy of 1 corporation so I took matters in my own hands and run Linux: it rarely fails. When it does, it's usually my fault, so I can assign blame, learn my lesson and move on. Otherwise, I'd have to accept that every so often, a proprietary vendor update may brick the device.
Sometimes while reading an article I silently curse the load time or layout or terminology. Then I see others here discussing that.
HN does not disappoint.
From a personal perspective, the discussion quality on HN (on average) has declined over the past couple years. In the past the majority of comments were more substantial; it was more common to see paragraphs of well thought-out comments that delved into nitty gritty details when addressing a point.
But in the past couple years there's been a gradual increase in single-sentence comments that contribute nothing to discussion but only end up taking up screen space. Like the random person who wants to throw in their 2 cents, but have zero cents to offer.
Even HN is not immune to Eternal September.
Complaints about HN getting more like Reddit go back to 2007, before it was even called Hacker News.
Short articles don't have much to discuss. There's no need to get mad that someone isn't laser-focused on the two sentences you think are most important.
> Bravo.
'Bravi', or 'brave'(f) would be the plural of that.
This thread is almost as shit as Catalina.
I have never heard anyone but you say this ergo it is wrong.
People go into Karen mode here, yes they bought expensive piece of hardware with expensive operating system. Though site is rather technically minded they still underestimate how complex are those things.
It just another "Dropbox? Who needs that I can do the same in an hour", but somehow they don't realize when you have a budget, timelines and as always not enough developers who understand this specific thing - it is different than sitting on your own and making perfect thing. (which then if scrutinized by some other dev would be labeled as crap :D)
"Whenever the band found brown M&M's candies backstage, they immediately did a complete line check, inspecting every aspect of the sound, lighting and stage setup to make sure it was perfect. David Lee Roth would also trash the band’s dressing room to prove a point -- reinforcing his reputation in the process."
So, in summary: the team of roadies DIDN'T always do a a complete line check. And DLR would just make life hard for the hotel staff. That's just dumb all around.
Most content is also available on non-spammy, ethical websites, and we shouldn't support bad actors.
Well, this is not nitpicking. The word "bricked" has a meaning, and this is not what the article describes. From the title, I thought that this update was actually bricking the computer. Thanks to the commenters for pointing it wasn't the case.
Regardless of what you think about HN comments, this is antisocial behavior, and can't possibly raise the quality of discussion.
Also I’ve read that submissions with too much early comment activity get removed automatically for seemingly being too controversial. If so then maybe that rule should apply to comments as well. Controversial comments could automatically sink lower despite upvote count, as well as older comments.
"What is life?" "What is good?" "What is need?" "What is you?" "What is what?"
And those discussions will never end unless the time itself ends.
> Crush your enemies, see them driven before you, and hear the lamentation of the women!
(this is a philosophical joke)
1. their devices are (or are on a path of becoming) non-general purpose computing devices.
2. developers are treated like junk on the app store.
Developers are forced to operate in a "market" dominated by a few oligopolists that are obsessed with control, both on their own behalf and on the behalf of governments.
Bravo.
No problems, still bored.
And I’m not talking hearsay here, I own one and the keyboard is as shitty as the entire internet says. Comparing to my previous MacBooks that had no keyboard problems when used outdoors too.