I've been having some sort of severe performance issue in one form or another since Mojave/Catalina or so with a 2017 MBP and a 2020 M1 Mac Mini.
The symptoms is always generally poor performance after the system has been running a while (4h to a week, varies), usually with WindowServer using CPU cycles non-stop and UI that felt choppy across all programs.
This seemed to happen frequently after "opening many files", like doing some recompiling with Xcode for a few hours, or indexing a large volume with Spotlight. Rebooting helps temporarily.
Today I realized that data read/written since boot was about 1TB in a few hours on a brand new OS install, and I traced this back to the com.apple.Safari.History process. Somehow having bookmarks and previously using Safari 15.x caused a huge amount of I/O that wouldn't stop - the solution was to remove all bookmarks and reading list items. Performance was immediately back to normal, no reboot needed.
So just logging in with your iCloud id, you could be "importing" whatever performance problem you're having on a new install.
I recommend you reboot and take a look at your disk I/O stats - maybe this will help someone!
Yes, noticed this some time ago with WindowServer taking all my CPU usage but only twice so far for this year I think so it was okay. When it occurs every day then I will start to worry then
IIRC, when the M1 first came out, there were a bunch of people saying their SSDs were being worn out super quickly, citing SMART statistics. Perhaps this safari process was the culprit?
BTW, Apple totally lied when they said they "fixed it" in an update and it was only a "reporting issue". It's not fixed, and it was absolutely f'ing not a reporting issue. People's SSDs have already failed because of this, and obviously they're soldered.
> People's SSDs have already failed because of this, and obviously they're soldered.
Any links to more info about this? I’ve had my eye on a 16” Pro with an M1 Max for a couple weeks now but want to make sure it won’t have issues like this.
wow Safari is the IE5 of browsers now, having to do so many work arounds like we used to have to do with IE5 and now its killing hardware that you can't even replace.
It’s only “behind” relative to Chrome, the dominant browser which more or less sets the standards these days. Calling Safari “behind” is judging browsers by how similar to Chrome they are.
Also Firefox and even Edge when the IE version was being developed. They've also refused to implement features which would bring web apps in line with the the capabilities of native iOS apps (things like push notifications, offline support with data that doesn't get cleared every 7 days).
Yes I've seen something similar. WindowServer taking up all the memory.
In one case I also had it crash while I was out, with every open program opened several times on the dock. I didn't do this and in fact macOS doesn't let you do this. Everything was hanging completely so I had to turn it off and on again :)
But this kind of thing does not instill a lot of confidence
Do you run your laptop on non—native resolution? Or an external monitor with a non-native resolution?
I'm runnign on 2018 MBP, 16GB RAM + 4k external monitor. I experienced the same type of issues a couple of months back — high WindowServer CPU, mega choppy UI after a few days of use. Initially thought it was Safari, but it kept happening with other browsers as well. Researched it a bit and found a thread where someone suggested running both the laptop and the monitor on native resolution. Haven't had any problems since doing that. WindowServer sits at about 10% CPU and ~2GB RAM, current uptime 10 days.
I think I had the same problem as you (similar hw and setup). Do you have a discrete GPU? If so check the link below out. For me, this behavior was due to OSX flip-flopping between discrete and internal GPU. Once I set it to discrete all the time (I'm always plugged in) the problem went away.
Seems to me a different issue than what people are complaining about atm... but what do i know, might be one and the same.
Edit: very weird... Looks like this setting was reverted for me. I just updated from 11.? to 12.0.1, so I wonder if the installer undid some of my changes. No performance issues, but I've literally just my computer on for the first time after the upgrade.
I'm also pretty much always plugged in. It's likely that the issue was manifesting before switching to a 4k monitor, but after switching the issue became unbearable. E.g I'd wait 2-3s to switch between windows or desktops. Running both screens on native resolutions and I'm switching between windows near-instantly.
:( Sorry to hear that... I know it must be so incredibly frustrating!
Sounds like it isn't an issue of flip/flopping between discrete and integrated GPU then... must just be the integrated GPU and its ability to deal with non-native resolutions.
SafaribookmarkSync, Cloudd, along with Content Cache has been problematic for years. Especially if you have huge number of Tabs and Bookmarks that is being synced across devices.
Along with iCloud syncing for one reason or another have relatively higher probability of being messed up during update, and you end up with a scenario where something is to probably synced and it keeps trying it over and over again.
I just ordered the M1 Max with 64GB RAM because I'm constantly getting the "Your system has run out of application memory." popup while working on Lunar[1] in Xcode.
Every time this happens, Xcode uses about 4GB of RAM (probably because of the monolithic UI storyboard of Lunar) but it should still leave enough memory for my other non-memory hungry apps.
But then I open Activity Monitor and I see WindowServer using ~80GB of memory [2]
The only remedy is either `killall WindowServer` or a full reboot.
I've been using an M1 MacBook Pro and Monterey since the first developer beta but this only became an issue in the last 2 months or so.
It is virtual memory, the real memory size is usually sub 1GB. But I just can’t find any other culprit, and killing WindowServer is the only thing that helps.
I have no idea how to troubleshoot this and don’t really have time for it when working on hotfixes, I just want to get back in a state where I’m not facing an imminent system lock up where I can’t even save my recent work.
Do you tint your mouse cursor (new Monterey accessibility setting)? I noticed this in the last handful of developer betas, and it definitely shipped out in 12.0.1: https://bugzilla.mozilla.org/show_bug.cgi?id=1735345
Although that particular issue will show in application footprint, not WindowServer. I see WindowServer buildup on one of my M1s still, even in 12.1 beta :(
Activity Monitor shows a "memory footprint", rather than just a simple sum of virtual memory usage. Among other things, this number splits memory usage for resources shared across multiple processes (e.g. system libraries) and attributes a fraction to each of them.
I too have the problem with WindowServer although for me it seems to happen with just regular usage. I have not rebooted in 48 days, and currently it is using 10GB RAM (I have a MacBook Air with 16GB). I have seen some people claim it has to do with using a display scaling setting other than the default (I use 1280x800).
Yes, I agree, but it is a memory leak. It doesn’t START at 10GB. Also closing windows does not help much at all.
To answer your question, I have a total of 73 windows open at the moment. 11 of them are from applications with a single window open, 8 are from iTerm, and 54 are from Sublime Text. I am aware that is quite a lot for Sublime, but that is just how I use it.
Regardless, I just quit Sublime Text, and the memory usage only dropped to 8.9GB still absurdly high for having 19 windows open.
Hmm, that is interesting. You have lots of unaccounted for memory, which is not normal at all. Would you happen to have the full footprint output? Trying to see how the 200K individual VM_ALLOCATE regions are distributed in side.
I have been seeing all of the above problems on my 16GB M1 MacBook Air for probably at least 6 months now.
WindowServer has consistently been the main culprit but occasionally other apps like to chip in where they can too. Often that means there isn’t one single process which I can kill to reclaim enough memory to continue working without being promoted again or having MacOS forcibly close all my applications on me. As a result, reboots appear to be the only real solution for me.
Interestingly though, in the past few weeks I have been working with similar numbers of open Sublime Text windows as yourself and I can confirm that since then, ST4 has been a regular and serious offender when it comes to memory usage in Activity Monitor.
Great to see you pop up on HN. I discovered Lunar a couple of weeks ago after wondering whether something like it existed and it's been a great tool so far. Solves one thing and solves it well. Keep up the great work!
If you use say Proxmox and create your own macOS image and compile your own Opencore then you should be OK. You can do same on VirtualBox. (On a PC, cause I doubt older MBPs have the necessary stuff like VT-d.)
Of course, once Apple kills off x86-64 macOS it is going to become more complicated.
That won't be for a very long time. There's a lot of Mac Pros out there. They're still supporting the 8-year old 2013 Mac Pro, the trashcan one. They sold a lot of the revised Mac Pro to animation studios and such. I wouldn't count on support for those being dropped for at least ten years.
That seems like another symptom of the same problem.
Apple locked me into their ecosystem...but that is their right, after all it is their platform.
Apple sold me faulty hardware/software but I have to buy more Apple products (new hardware, support) since I am locked in and cannot do my work on non-Apple hardware.
Their behavior may not be illegal (IANAL) but I find it quite user hostile.
The main problem I see with all this is that since Apple is getting away with it, more and more companies are trying the same tactics and normalizing user lock-in and charging extra to fix faulty hardware and software.
As someone who hasn't been in the Mac ecosystem since his and his mom's iMac both broke in the same way in high school, how does the Mac ecosystem lock you in?
As far as I know you need to use Xcode to distribute the software, or to write macOS-native software. I have written plenty of Java Swing apps without touching Xcode at all.
You can use alternative toolchains like like cctools-port[1] and xcbuild[2]. See, for example, this article [3] on how to build Swift UI apps for iOS using Linux.
It's very weird to be talking about "windows, mac or linux" developers.
Desktop application development today is maybe about 1-5% of all development, while overwhelming majority of backend development - the largest category - targets Linux.
Percentage of the way to being as good as all desktop operating systems ought, at a minimum, to be:
Windows: 70%
Linux: 20-80% depending on your hardware, what you're doing with it, and how lucky you got with the current kernel & libraries your distro is running.
MacOS: 90%
So when Apple fucks up really bad and MacOS temporarily drops to 70%, no, I still don't switch. Every desktop operating system is shitty, but there are degrees of shitty.
Probably because other brands don’t do anything novel to warrant such loyalty. Apples hardware and software stack is undeniably unique which in turn makes it a brand that people will put up with issues because well, it has things other systems don’t.
These issues notwithstanding, the total user experience of owning a Mac laptop is far superior to any Linux or Windows machine I've tried, at least in the past 3 years or so.
Two concrete examples: trackpad feels like shit on both Windows and Linux, installing or updating drivers and some apps on Windows constantly switches back to legacy modes to display the installers (with different fonts, no subpixel smoothing etc.). If it's not a deal breaker, it is annoying as hell and something that almost never happens on my Mac.
yea it takes time but eventually open software will be better, just look at how well Blender is doing and even Darktable is way way better then Adobe Lightroom can't understand why anyone uses that anymore while a free alternative that is better is out there.
^ This. It's not that I don't want to switch to a different company, it's just everything else is a worse, overall experience.
I'm on Windows every week or so, and while it feels much faster than Mac, on the same hardware, there are just too many little things that annoy me. (have tried a few Linux flavours too, not really a contender for my use case)
OP was describing very high-spec hardware which locks up and becomes unresponsive under their normal workload, potentially losing work or wasting their time. That seems like a much more significant issue than inconsistent font display or the feel of the input device.
Your experience must differ, but if I spent $2000+ on a device only to find that was locking up (due to poor memory management in a standard component!) under my standard workload, I would be livid. I would probably return it, and then tell the next 10 people I meet about my poor experience.
That said, I have run across a Windows 7 bug where the font manager would repeatedly fail and leave its cache/log on disk, only to restart and repeat until the hard disk was just about full. Complaints about the poor state of software engineering as a discipline go back to the beginning.
As a counterpoint, I was a Linux user for ~20 years (eventually with a macOS dev machine on my desk as well).
I got so sick of troubleshooting issues myself, I switched to Windows 7 and have never had to troubleshoot my OS since then. I don't google Windows issues, I don't have crashes, nothing.
It's nice to have a machine that I can just work on and not constantly debug.
(I 100% expect privacy advocates to admonish me for this, but I use hardware I assembled myself and block MS telemetry.)
In a world without MacOS, I'd probably be running Linux on a really beefy Windows machine under virtualization, for work. Desktop Linux is so much more stable and pleasant to use in a VM with Windows or MacOS handling the actual hardware. Still a bit unstable, but not nearly as bad.
> These issues notwithstanding, the total user experience of owning a Mac laptop is far superior to any Linux or Windows machine I've tried, at least in the past 3 years or so.
That.
I don't love Apple. They're a corporation, so at best our interests temporarily align. Loving them would be absurd. I really really want them to have strong competitors. I'd love to be back on an open source OS full-time, like I was for many years.
Unfortunately, they're so far ahead of the competition, that they screw up, sometimes even in a few ways at the same time, and people come out saying "LOL and Apple fanboys won't switch to Linux even now", and that's true... but it's because I'd be trading a few problems for a few score problems.
I ran Linux on laptops and desktops for about a decade, as my main computers. Ubuntu near the end, Gentoo for about five years, Mandrake really early on, a little time with Fedora somewhere in there, a sprinkling of Debian. I still try it out every year or two! I wish it weren't, but it's still much, much worse, and in the best case slows me down and gets in my way more than a "bad" Apple machine with a "bad" version of an OS X / MacOS, barring actual faulty hardware.
Tesla. So many people seem to have problems with the cars/lack of quality only to say “still love the car” or “it’s my xxx number Tesla and they’re still doing [problem]”
It’s really something, isn’t it? I’ve never regretted my decision to switch from macOS to Linux over a decade ago. Having to restart a 64GB machine constantly because it runs out of memory? Meanwhile my 4GB Linux computer shows a 38-day uptime (only because I accidentally unplugged it last month) and my two Linux servers have uptimes of over a year.
Just to be clear: This perfectly fine experience you describe is also exactly the experience of 99% of Apple users, including me. You only hear the ones that complain.
The ones that don’t complain have low standards. When I switched, although OSX was pretty solid, it was trading frequently spinning beachballs for the first consistently responsive computer I’d ever used. I had come to regard macOS sluggishness as normal, as do most Mac users, because they haven’t experienced anything better.
Is this sort of sweeping generalization really necessary? I can tell you that it's not true in any case. I'm one of those Mac users that doesn't complain about performance, because I rarely see the spinning beach ball.
I haven't seen a spinning beachball on MacOS in years. I'm not even sure it still exists anymore. But also only rarely have a sluggish system, with or without a spinning beachball.
That doesn't mean linux is any better. I built a brand new ryzen x5800 machine late last year, running linux mint. When I first built it, the computer has had duplicate unremovable mouse cursors appearing on the screen. And my trackpad sometimes randomly became unresponsive (or one time, the up and down directions flipped!). And neither suspending nor hybernating worked - the machine wouldn't come back up.
All these problems have since gone away thanks to software updates and (importantly) a BIOS update. It turned out my USB controller wasn't being configured correctly by the motherboard. If I didn't happen to check for BIOS updates, I'd still have lots of problems.
And that was still a better experience than the HP laptop I bought a couple of years ago. That laptop had so much HP crap preinstalled that it sat on 30% CPU usage at all times out of the box. The biggest offender was some HP process doing microphone noise reduction. It ran all the time regardless of whether or not the microphone was being used - and utterly demolished the machine's battery life.
Apple is far from perfect, but they have a better reputation than most vendors for a reason. And their technical support is are excellent. I've had thousands of dollars of free hardware repairs & replacements over the years when things have broken in my apple machines. I wish they didn't break so much; and I wish they prioritised fixing software bugs over adding features internally. (APFS aside, I would trade the last 5 years of macOS features for better stability and performance in a heartbeat.) But there's a reason they're so well liked on HN. They're still pretty excellent in comparison to the competition.
Huh? What point are you trying to make? I don't care how well macOS runs on my machine. I care how well any given computer gets out of the way and lets me get my work done. Today that means intellij, the rust compiler, firefox and nodejs. I want to spend my day working. Not futzing around with HP bloatware. Not rebooting due to apple memory leaks. And not guessing why my keyboard needs to be reconnected.
In your story, you built a machine, and Linux didn’t work well on it at first. Later, it worked fine. Meanwhile, macOS never did and probably never will run on it. And you introduce this story as an illustration of how “Linux isn’t any better.” So, to borrow your words: Huh? What point are you trying to make?
In my story linux didn't work well at first. Then with some time and work, it worked fine. An HP windows laptop didn't work well at first - but after I spent some time removing the HP bloatware, it worked "fine". This article says that mac laptops aren't working well because of memory leaks. But in a few months Apple will probably fix their software bugs and they'll work fine too.
You said:
> The [mac users] that don’t complain have low standards
I agree with you. I'd go even further and say most users have low standards, because the out-of-the-box experience with most modern computers is pretty bad. Apple users should absolutely complain more when the out-of-the-box experience with their computer isn't perfect. We all should.
Apple customers keep buying apple computers despite their issues not because we have low standards. We do it because the alternatives are even worse.
As an avid linux user I can tell you. Linux users aren't "trained" to have low expectations. They are trained to be masochists that enjoy pain. The raw amount of configuration, fixes and setup work that has to go into making some linux distros work is astronomical.
If nonlinux users are trained to have low expectations, Linux users are trained to rape themselves constantly.
I see your point of view a lot but I don't think configuring a system is that hard and I think it's preferable, compared to have tons of complexity preconfigured without your knowledge.
Here's my experience with Arch Linux:
1) You install what you need to install
2) You read the doc and configure to your liking, learning what the system can and cannot do and how it works with everything else
3) Things generally work as documented and you go your merry way
4) When something break or doesn't work, you have the knowledge or a general idea of what to fix
Compared to proprietary systems where there is no documentation and little configuration, when things break or misbehave you're essentially screwed. Will you wait for an official fix or look online for reverse engineers?
You bought a black box and you can't just open it and fix it.
I'd rather spend a day learning about my system during the installation and having it work for years, instead of buying yet another black box I can't open.
>I'd rather spend a day learning about my system during the installation and having it work for years, instead of buying yet another black box I can't open.
Learning linux doesn't take a day. Second off those black boxes statistically are more reliable than linux EVEN when you account for all the things you talked about.
The reason for this is simple. Microsoft has a business advantage and unfairly strong arms hardware manufacturers to make their stuff work with windows. This unfair advantage makes the windows user experience better and more reliable than GNU linux whos' developers constantly have to play catch up with hardware.
In the last 20 years PC hardware has become ridiculously faster - and yet the user experience seems to have changed little.
I saw an article ages ago whereby Microsoft apparently had a dependencies approval gateway. If you made any change based on Windows you had to remove/reduce your dependencies or make a good case for not doing so.
The point he's trying to make is linux on a Ryzen isn't as good as MacOS on an M1. The user experience is inferior. That is his point and his point is fact. Hope you understand now.
user experience is inferior? how so? what nonsense, i have heard countless stories on dumb things Apple forces on its users while on Linux you can just use Vulkan... Imagine if your app can't contact Apple servers they don't launch or take minutes to open...and when they do work they are recording information... why on earth do people still use closed source software. Now we have to deal with Safari crap... like we used to deal with IE5, having to have workarounds for browser standards Apple refuses to follow or implement.
Don't waste your breath, let them be 'happy' in their little cult.
I think Apple users are often mad because subconsciously they know they are taken a fool.
It's not a cult. I use all three operating systems and I'm highly proficient using linux. My distro of choice is Manjaro and NixOS and I have a lot of experience with debian and ubuntu as well.
Each operating system has benefits and downsides. For user experience linux is definitively the worst. Definitively. The differences are so large that it's almost comical how someone on this planet can't see the difference. If you can't see this then you are the one that is part of a cult.
- Hidpi support is spotty (its getting better but still a bit spotty)
- Smooth scrolling in applications is inconsistent at best
- The keyboard shortcuts for moving the text cursor around are inconsistent between programs
- Hardware support is much more of a crap shoot. Linux has a harder job than macos in trying to support every combination of janky hardware out there. But as a user, I don't really care. I just know that if I buy a mac, the OS will work perfectly with the hardware on offer. That isn't true on linux.
- Lots of useful software isn't available on linux. Eg, I love Monodraw, but thats macos only.
- App distribution on Linux is a mess. Apt? RPM? Snap? Flatpack? Etc etc. I have 2 copies of discord installed for some reason, and I have no idea what the difference is between them.
Etcetera..
Linux gives you the choice and freedom to spend an unlimited number of hours customizing everything. With linux I'm in complete control and I love that. On macos, things usually just work out of the box. I love that too.
No OS is perfect. There's tradeoffs with everything. If you don't understand other people's preferences, that doesn't make you right. It makes you ignorant.
He's obviously comparing a linux PC experience with an apple OSX experience. It's a completely valid comparison but you're just being mean an inappropriate.
Tell that to my tech lead who spent more than 8k buying and repairing his macbook due to constant motherboard issues.
I love macs but don't dare to try them due to soooo many people in my circles seem to have very frequent HW issues with them. It seems to be either in the shop being repaired or having issues with connecting bluetooth stuff.
I'm used to Thinkpad reliability so not looking forward to trade that up for some pretty pixels.
My lenovo laptops had 0 issues in the past ten years. Zero.
But now I'm web development and I need a mac for Safari to test locally.
I manage a fleet of macs. I have not seen any issues like that. I'd ask what your friends are doing to their computers to be seeing such a disproportionate amount of hardware issues.
> The ones that don’t complain have low standards.
Wow, well, thanks a lot. I've been developing on both the Mac and Linux for years and I don't have this issue on any of the multiple MBPs I use, including my personal 2013 model. I'm not saying I have no problems on the Mac, but they're not worse than the issues I had with Linux, especially on laptops.
I switched to a 13" M1 after having a Dell XPS 15 as my daily driver. I had high standards and Apple blew them out of the water. I see some beach ball on occasion, but I slam the machine hard. (8gb ram). It has been more reliable than the XPS 15.
Better than a Dell XPS 15 is no standard at all. The Dell XPS 15 is an utterly terrible laptop that breaks if you close the lid and put it in your bag when you get up from your seat. It's a portable desktop falsely advertised as a laptop due to its many flaws. If the best that can be said for apple is better than Dell XPS 15, that is a really, really damning review. I hope it's vastly better than that for you.
> Better than a Dell XPS 15 is no standard at all.
so what is the laptop that would meet this standard for which you're making a judgement?
So far, apple hardware has been superior to any other manufacturer, even with their warts, and their anti-consumer practises (of not having replaceable batteries, etc).
Not OP, but I have to chime in here- Dell's XPS line somehow got a reputation that it absolutely did not deserve as a good laptop to run linux on (or a good laptop at all).
I've been perfectly happy with a surface book 2, and various Lenovo and HP laptops. They each have better and worse product lines. Currently rocking an HP elitebook for personal use and a MB Pro for work.
The hardware quality is more or less the same on both. With the pro, you get a moderately better touchpad, a bit thinner and metal chassis. On the other hand, crap keyboard and the abomination that is the touch bar, and no legacy ports.
Personally, I don't think I would buy Apple for myself ever again, but that's just me.
The fact that we even have to debate about which laptop is good for Linux or not makes me not want to bother. I switched to Mac 12 years ago, and I never had to bother thinking about choosing hardware since. It just works.
No one will ever convince me that a laptop with a web cam whose position is limited to looking straight up your nose is quality.
I primarily navigate by keyboard, so the difference between an acceptable and amazing trackpad is pretty minimal for my preferences. On the other hand, port wonkiness, temperamental wifi, okay keys, mediocre battery, driver issues and an utterly useless webcam convinced me to ditch the XPS the first chance I got. It's amazing I used it nearly 18 months.
>So far, apple hardware has been superior to any other manufacturer, even with their warts, and their anti-consumer practises (of not having replaceable batteries, etc).
Milage may vary with other manufacturers, it doesn't look like you would change to another manufacturer. Which is ok.
Check business notebooks, from HP or Lenovo. (I forgot one particular Dell series.)
Ok... so what's the laptop that can compete with the new 14 and 16 Macbook Pros then and doesn't have some stupid flaw? The issue with waking up in bags seems to be caused by Modern Standby which is on most new Windows laptops?
I would go with a Lenovo or HP machine before a Dell. Even if they all have the sleep issue (I don't know if that's the case), both Lenovo and HP have better build quality.
I've had multiple Asus laptops (UX305, UX330 and most recently S435UA) and all of them ran Fedora flawlessly out of the box. Granted, none of them had Nvidia GPUs which is the usual pain point of desktop Linux.
I have had the opposite experience. Switched from a MB Pro that kept getting progressively worse over the course of a year (keys giving, overheating, lock ups) requiring rather annoying applecare interactions to an XPS 15. Granted it is not quite as pretty look at, but blows the mbpro out of the water performance and stability wise.
> The ones that don’t complain have low standards.
> they haven’t experienced anything better.
I don't want to gratify this pointless flame any more than to say that I have used all of Windows (XP, Vista, 8, 8.1, 10), Debian, Fedora, Ubuntu, and macOS as my daily driver at one point or another, and I currently use macOS because I have yet to experience anything better. There is a simple explanation for your problems here: something is wrong with your system (or you are overtaxing it) if it beachballs constantly. This is not typical macOS behavior, especially not since they switched to SSDs. I actually haven’t seen a beachball in years.
More like "specific enough" instead of "high enough". Use Apple all you want but don't lie to yourself by saying Windows isn't a better, more powerful, more stable operating system.
It just isn't as polished, and even that's debatable.
You are confusing your own personal experience for the standard experience of macOS users. Personally, beachballs are so rare for me in the past few years that I'm always amazed when I see one.
This has been my experience as well. I've had an M1 MacBook Pro since they came out and I've never had to reboot my machine outside of software updates that required me to. It has always just run.
I use my computer to write Elixir code, write music with Logic and Studio One, and edit photos/videos. Maybe none of those tasks trigger the problems other people are having.
I don’t know what you’re doing on a 4GB machine but I promise most workloads that create significant memory pressure would start creating problems for you, just like on macOS.
I use a dual xeon, 64GB Linux desktop which performs reasonably well.
A colleague runs a Linux machine with 16GB of ram. Compiling the same project can cause OOM issues for him and the kernel starts whacking processes.
Even if he closes all non-essential processes, his builds run 10x as long as mine due to memory constraints.
My main point here - let’s not pretend your 4GB machine is being used to do any intense local work.
The inverse of this is seen on my Intel MacBook Air and my colleague’s M1 MacBook Air. The compilation time for our code is 35 minutes or more on my machine and only 3 minutes on the M1.
> My main point here - let’s not pretend your 4GB machine is being used to do any intense local work.
That depends entirely on what you consider intense local work. Just because you are not compiling doesn’t mean running your 15 microservice stack locally is not intense.
Memory intensive work wasn't the topic of conversation. GGP's comment was about how often he had to reboot his laptop despite it's immense memory budget, not about how much faster it was at performing memory-intensive tasks.
4 GB is maybe small nowadays, but memory-intensive is relative.
I'd argue that _most_ programmers or even data analysts don't need more than 8 GB to perform their actual work.
But realistically they need 16+ anyway because of opening 200 browser tabs, memory-heavy IDEs, etc. 32 or 64 is nice for sure, but definitely not needed for actual "work" workloads for the most part.
Of course it will run out of memory if it's taking more than 16GB of ram.
Most Linux distro suggest you to setup swap so instead of crashing it could use the swap.
I personally don't use swap, because if something is consuming too much ram I most likely want it dead.
I don't know what are you compiling, but requiring more than 16GB of ram for compiling something sounds lazy and unneeded.
Likewise you can use a browser extension which reclaims ram from unused tabs.
I can understand the need to have 64gb to do video editing or rendering, but most of the times we need that much ram is because our software is badly written and we don't want to face the problem, not because of the complexity of what we're doing.
I never regretted moving from Linux to Mac also a decade ago. Mac has been fairly smooth for me, still on a 2013 mbp that just needed a battery replacement. Linux, however, was pain, quite often.
its a lot different now, unless your running it on proprietary hardware without official driver support, like the random stuff laptops have these days, but now we have many companies making laptops for Linux so it has improved a lot. Why give money to a company to use their software when you can use open software and help fix and improve something everyone can use for free.
Maybe not constantly but I have to restart my 64 GB Linux machine because it runs out of memory. Linux's OOM killer is terrible. I literally have to wait half an hour while my desktop is frozen for the OOM killer to kick in. And no, spamming the key combo for invoking the OOM killer once this happens doesn't help.
Use earlyoom or another oom daemon for a 1000x better experience and consider figuring out what is leaking. Do you use gnome? It used to have a bad leak that was supposed to have been fixed.
I think it's mostly firefox just holding things in memory since so much is free. Eventually it just all adds up and eventually I try to compile something and then it will run out of memory.
Yes, I can go setup an oom daemon, but Linux should handle this for me out of the box (ideally with some sort of logic to avoid killing something like X taking down every single program I have open). Getting stuck for half an hour or longer is just not acceptable to let happen.
While the tooling and system stability are better, the desktop and app experiences are far worse. It’s buggy, half-baked, and crash.
I regularly hop distributions and do clean installs to see if that will fix it but they’re all the same — pop!_os might be slightly more stable in that regard.
I’ve been using Linux since 1992. I’ve had several year periods in the 90s and early 2000s where Linux was all I used.
I can ditch GNOME, KDE, and XFCE and go back to twm or i3 or fvwm or Window Maker, but I’ve moved beyond thoses experiences. I want a more functional desktop.
But, I see now why I paid more to put up with Apple.
Can you please explain what exactly you get from the Apple desktop you don't with any of those options? I'm currently forced onto a mac for work and I find it lacking in almost every way other than an ever so slight advantage in beauty.
So besides dpi-scaling issues linux is known for, whats the draw? Some hotkey that does a thing? Some visual candy? Smoothness in transitions?
Whatever it is, it's hard for me to imagine it being worth the tradeoff. I can't even name desktop workspaces on a Mac!
I'm reviewing the "new features" section on Monterey: https://www.apple.com/macos/monterey/features/ and I'm not seeing much other than a few known areas (multi-device support if you bought fully into the apple ecosystem, etc).
With most "I tried linux desktop" people it's a problem of DE/WM, but you seem to be beyond that so I'm genuinely curious.
> Can you please explain what exactly you get from the Apple desktop you don't with any of those options?
A more cohesive experience across the desktop and applications, eg.:
- The look and feel is uniform. Admittedly somewhat less so, these days now that the Apple Human Interface Guidelines are largely ignored.
- Mac apps and generally more stable.
- Mac apps generally have more features.
- The Mac apps I use generally tend to tie into the hardware (hw acceleration, drivers, etc) much better. The benefit of controlling the vertical, I suppose.
- The keyboard shortcuts both exist and are the same across apps.
- The design patterns.
- The AppleScript/x-callback-url, etc).
- Third party apps generally tend to fit well into the above.
On Linux, almost everybody seems to have their own ideas on how an app should be arranged. With full desktop environments, like GNOME and KDE lessen this, but it's still really, really common.
That said, my money is still on Linux and other open source OSes, because I don't want a single company telling me what I can and can't do.
Not OP but I have extensive experience as a user of all three major desktop systems. Aqua, the macOS desktop, is by far the most stable and consistent, and in my experience the most usable and powerful as well.
A few things that stand out about it to me vs GNOME 3, KDE Plasma, Xfce, LXDE, and whatever the Windows interface is called:
- The consistency of the menu system, and being able to search and use the menus of any app from the keyboard with shift-cmd-?. This is like having Emacs M-x or Sublime/VSCode shift-cmd-P in every desktop app.
- Being able to assign custom keyboard shortcuts for any menu option in any app.
- Emacs keybindings for editing text in every text field (including on the web). I believe GNOME Tweaks is supposed to do this, but I could never get it to work reliably and universally like macOS.
- Native app ecosystem. Third-party Mac software is generally the most polished, though not quite as much exists as for Windows.
- System animations. This is a small one, but it makes things more fun and makes the whole system feel fluid and “organic,” for lack of a better term.
Linux works fine for me, but I don’t use a desktop. I totally agree with you that they are disasters, with poor documentation and multiple conflicting ways to make settings.
Firefox crashing is probably a Firefox problem.
You can set the scaling to be anything you want on X, just set the Xft.dpi number in your .Xresources file.
One problem with your Hackintosh install is that it’s closed source, no? As far as you know it’s exfiltrating your financial information to China.
macOS still crashes, has weird bugs and every 6 months they go and change a bunch of shit that either deprecates functionality, causes conflict with how things work or takes another handful of your privacy.
Grass isn't always greener.
Edit for an example, I'm stuck on Catalina because they tried deprecating functionality my firewall uses. And the window shifter shortcut key program I use won't work in Big Sur.
> and every 6 months they go and change a bunch of shit that either deprecates functionality, causes conflict with how things work or takes another handful of your privacy.
No, that's Microsoft Windows. Apple only does it annually.
Windows has some of the best backwards compatibility around. Stuff from the 90s still works on Windows. Meanwhile macOS deprecated all 32 bit programs.
The kinds of things the post above complained about breaking on macOS—third-party software that messes with system internals in unsupported ways—also break frequently on Windows. To say nothing of the first-party breakage of Microsoft constantly moving around settings, re-enabling or re-installing annoyances you turned off, and removing your ability to easily get rid of them, and adding new ones behind your back.
Microsoft really only tries to preserve basic compatibility for well-behaved applications that aren't tightly integrated with any OS components or specific hardware/drivers and aren't doing anything that Microsoft disapproves of. Anything that strays outside those boundaries will run into trouble, and a lot of software ends up falling outside those boundaries even if it didn't really need to.
Both operating systems are pretty bad at letting you wield control over your own computer. But macOS doesn't try as hard to obscure the BS or obstruct your attempts to tame it, and macOS actually lets you refuse updates that you don't want.
Yes and many of the major Linux distros have discontinued or plan to discontinue official 32-bit/multilib support. Why? A lot of 32-bit libraries have serious vulnerabilities that no one is fixing. Essentially most of the world has transitioned to 64-bit except for a few key areas that are still trying to run libraries that are decades old.
Sure, macOS still crashes, I didn't claim it doesn't. But macOS crashes with a lot less regularity for me.
I use Linux as my daily driver and the apps I use (GNOME Terminal, other terms, Spotify, Slack, Brave and Firefox, and the GNOME desktop itself) crash multiple, sometimes several times a day on my two (work, personal) machines.
This happens across diverse machines, as well as across multiple distributions (though again, pop!_os, so far seems much less trashy for whatever reason).
what are you doing to cause these crashes or what hardware?? i use manjaro and pop_o, makes no sense since i get less crashes then windows and my computer is on for days without anything crashing, other then me writing a bug and crashing my own stuff :) i do lots of work and gaming on them so i don't know why your experience is so different.
Desktop Linux is easily the least-stable OS I use with any regularity. It was my main OS from something like '01-'10, but after I finally gave OS X a try, and since Windows got its shit together some time late in the WinXP service pack cycle (or, arguably, Win2k, but that wouldn't run my games) and stopped crashing all the time, it's really hard to justify using desktop Linux.
Does the OS hard lock or completely crash? No... unless you have graphics drivers issues, which isn't unlikely. Then, oh man, yes, lots. X/Wayland crashes that restart the window server? Yep. Applications crashing pretty regularly or glitching out so badly they have to be restarted, including the basic applications distributed with the heavier DEs? Yep. And it turns out that your windowing environment crashing or the main program you're currently using crashing is really close to as bad as the whole machine blue-screening in Windows, from the perspective of the user. Using Linux makes me anxious, even though I very much know WTF I'm doing with it. MacOS and even (spits) Windows don't do that to me, any more. Now that I've experienced not feeling that way, I can't go back. I go years between work-lost crashes of any sort at all on MacOS. Linux, one month without such a thing would be miraculous.
If you build up from almost nothing and keep things very minimal and have very boring and stable old hardware, and do as much as possible from the command line, it can be kinda OK, but it's a lot of work to set something up like that, and ongoing effort every time you do something manually that'd be automatic or trivial on a more full-featured GUI desktop. If you start with something like Ubuntu or Fedora standard desktop installs, though, there's just too much that can go wrong, and it will, with some frequency.
I could tolerate some hardware or workflows not working and things generally being a little less convenient, maybe, if it were rock solid, but it's very far from that. The main problems seem to be that its entire graphics stack is incredibly fragile (Wayland doesn't seem to have done much, if anything, to fix that) and it's way too easy for a glitchy driver to screw up the whole system.
I've switched to Linux from Mac OS about 2 years ago, being frustrated by memory consumption, slowdowns and freezes. After trying a couple of distributions / DEs, I've settled on Manjaro (Arch based) and KDE. Delighted with flexibility, features and general performance and stability. While, before, I was constantly tempering with the environment and changing stuff and there was always "something missing", I've found myself not needing to touch configuration or change anything in my workflow for more than a year now...I did have some glitches and CPU usage issues when playing YT videos on Intel-based graphics, but have since been using Lenovo laptops with Ryzen 4000 and 5000, and it's been flawless without any tempering. P14s Gen 2 AMD (5850U + 32GB RAM) is the best laptop I have ever used (software development), and I change them A LOT...
AMD graphics on Linux nowadays are vastly more stable than anything else. NVIDIA is proprietary crap and Intel Graphics still have cross-platform bugs so often is almost comical.
The experience you are describing is more akin to what I remember being Linux in the '00s than nowadays. I run a machine with arguably a buggy Ryzen 1st gen motherboard (it's quite buggy on Windows too), and I have a very stable desktop experience with Arch Linux and KDE Plasma. The real game changer has been AMDGPU, I'm yet to have any sort of graphical issue with an RX 580 on Linux, and Plasma is arguably quite solid nowadays.
The only issue I had recently were audio issues which were both due to my buggy soundcard and a bug in Pipewire. Excluding that, I've had almost no issues with desktop Linux since 2017.
I think your experience is not universal. I would say windows has been far less reliable than linux on my thinkpad machine, but overall I have to say that I have yet to use an OS that never crashed on me, all software of this complexity is buggy.
Funny, I moved from Linux to Mac because I got so sick of no good laptops with full support for Linux and it's never ending subsystems for audio, webcams, trackpads, etc. And not like Linux UIs are known for being efficient and memory friendly either.
On the other hand my Linux computer struggles to go 10 days without a reboot due to the graphics driver bugging out, while my old MBA basically never needed rebooting.
Stop trying to spin it Apple's illegal behavior. The change was clearly unnecessary because once they rolled it back the phones kept on functioning just fine.
Modern hardware can and should last ten years. Accept no excuses.
I find the defense of Apple in this case absolutely fascinating.
You know how I can tell that Apple made that change not with the user in mind?
By the fact that until a 3rd party proved Apple was throttling older phones, Apple never publicly stated they were doing this. I’m fact, they vehemently denied it.
I saw this on a personal level, where my mother who took 2 trips abroad about a year apart, with the same phone (and she had a lot of disk space in both trips and she hardly installs any apps) found the phone an absolute pleasure to take pictures during the first trip, kept missing shots while waiting for the camera app to load in the second one.
And on taking the phone to the Genius Bar multiple times, she was gaslighted each time as the Genius showed her their diagnostic results which showed that no, her phone was in perfect health (minus a slight degradation in battery life).
Of course, a few months later the entire world learnt that, in fact, her phone had slowed down, and Apple was lying. What’s interesting is that Apple diagnostics tools (which should probably have a speed test?) also reported everything as being just fine.
Also: users were not warned about this "feature", many (most?) didn't know why their phones slowed down, and there was no way to disable the behaviour.
And I already mentioned in my initial comment that they didn't notify users. That's what the lawsuit in France was over, the lack of notice. I believe an OS update did eventually let users switch off the behaviour, and view their battery health.
why not just allow replacing batteries like phones did before?... seems like a dumb excuses to get people to buy a phone every year, make it impossible to replace the parts and slow down the phone. glue, solder, hardware lock parts etc etc until the phone is not even your property anymore.
If the goal was to increase product longevity why did Apple only admit this wonderful measure only after it was proven by a 3rd party?
If this was such an altruistic measure why would the geniuses at the Genius Bar deny that my mothers phone had slowed down and throw diagnostic tests at our faces? Why was the exasperated Genius, who could clearly tell that the phone was slow indeed but his successful diagnostics meant there was nothing he can do, reduced to swiping up all the apps on the phone to kill them, when we both knew that if iOS was working properly that should do nothing?
Apple has few and long-lived product identifiers; as such it is possible for specific problems to become well known.
It's extraordinary to find more than a handful of anecdotes about any particular model number of consumer Windows laptop that happens to be on sale at Best Buy today. There are so many, and they cycle so rapidly.
I want to buy a 16" M1 Max because it looks awesome...but my base-model 13" M1 is a beast of a machine and I use it every day for dev work. It has only 8gb of ram but it still chews through everything. And I got it for a great price brand new considering how much I've been using it since launch. In my own experience so far, if I buy that expensive 16" M1 Max it will be because I want it.
Note that I've had Dell XPS laptops that were configured with the same price point as a Macbook and while nice...those devices definitely didn't have the same attention to quality that I've experienced with Apple products. So it seems unfair to make such a broad criticism against Apple users at large!
Find me one laptop with a trackpad that works as well as a MacBook and we can talk.
I use a MacBook trackpad to make diagrams, for 3D modeling using Blender and FreeCAD, to draw schematics and route PCBs with KiCAD, and whatever else you’d otherwise do with a mouse, and it just doesn’t get in the way.
Over the years, I’ve been forced to use work supplied top-end Thinkpads, HP, and Dell laptops and none come even close to even the trackpad of my long retired, lowly 2009 MacBook Air 11”.
I don’t really care whether I use macOS, Linux or Windows: they all run the same applications. I also don’t care about the price of a laptop: what does it matter that a tool that I use day in day out for years costs $1000 more?
I care that it’s usable.
MacBooks are usable for things I use them for. The other laptops are not.
This right here, I got a Razer Blade from work last year, and found the touchpad so awful I assumed there must’ve been bad drivers installed. 3 driver installs later and I realized it just wasn’t as good as the MacBook trackpad, and I’d just need to pack a mouse everywhere I bring this laptop.
I love my 2017 mbp and I so want to tell every person posting workarounds for their $2000 brick please return it! Tim Apple won’t fix anything with all the $$$ in his eyes flowing in.
A finance professor once told me “if you don’t vote with your wallet you don’t get to complain”
Not a red herring, just confirming it wasn't a normal application memory leak (it's not, the malloc usage is fine–I think you're hitting the same issue as is mentioned in this article). The next step I would do is run sudo footprint $(pgrep WindowServer), optionally with -v to give you a really detailed list of why the footprint is so high.
> sluggish performance of a 2015 iMac since practically the day we bought it
I wish I had found this website when my MacBook Pro was still alive, maybe Chrome actually killed it after all. The low performance was so bad I moved to Windows 10 ("the last Windows system" they said), but after it notified my system can be upgraded to 11 I quickly moved to Ubuntu. I'm currently experiencing a few hiccups and reduced them with some memory tweaks, but still the experience is much better than Mac and Windows.
I feel like I'm a bizarro world, because I don't get this stuff: crashes, memory leaks, and so on.
I use: VScode, Idea, iTerm2 (with x86 brew), Logic (and lots of plugins), FCPX, Live (x86, and lots of plugins), Slack, Reeder, MS Office, Spotify, Chrome and other stuff. MBP M1 16GB.
What I do get, is some bluetooth issues (e.g. affecting the mouse cursor. If you start the bluetooth deadmon it fixes itself for a day or so.
You're not in a bizarro world. Most users are pretty happy with their M1 Macs (myself included) and HN just tends to upvote anything anti-apple even if it's an isolated experience.
I'd wager most people on this thread do not actually own an m1 Mac.
My experience (and that of others I know that have them) has been nothing but stellar.
Was it doing that before Monterey? I haven't upgraded to Monterey yet because I want to avoid any bugs that haven't been ironed out yet...so far my M1 8gb has been excellent and if those issues are introduced by Monterey then I may just delay upgraading until next year sometime.
Worth a read if you ever wonder why certain people on here are oddly fixated on the platforms other people are using. (Especially when they have no clue what you even do with your hardware.)
I’m generally happy with my Mac products, but I’ve also gone through long periods where my Mac would crash multiple times per week due to serious OS bugs.
Most recently, the Thunderbolt 3 issues were responsible for countless hours of lost time rebooting my laptop last year, and I’m not alone: https://rachelbythebay.com/w/2020/10/03/repro/
The weirdest part was that this occurred with only Apple products connected to my laptop. I couldn’t even blame it on 3rd-party products. It also worked perfectly fine before and after the range of affected MacOS versions. There were many reports of the same issue all over the internet, so I know I wasn’t alone.
Then one day Apple finally fixed the bug and everything was back to normal. But all those months of constant reboots and crashed and failed workarounds left a mark on my overall experience.
So while the Mac experience is generally good, it’s still far from perfect. Weirdly, I’ve had far fewer problems with my Windows system in recent years.
Claiming most people don't have M1 apple products is also bizarre considering almost all startups are using exclusively macs and we've all been waiting for the this release.
Its not like Apple doesn't release faulty hardware and software. Butterfly keyboard is a good example of faulty hardware and OSX Lion crashed all the time on the Macbook pros with discrete GPUs till Apple patched it.
I don't think it's really anti-Apple, but these articles do tend to take what seem likely to be very isolated events and present them as widespread, systemic issues.
I have absolutely no doubt that this person is having bad memory leaks (though their narrative about SoC memory or other elements seem unlikely to play any part whatsoever). Software has faults, and something in their environment or stack is causing problems. I have an M1 Mac and used every beta of Monterey, and now the release, and have had zero problems. Like nothing at all. I use XCode, IntelliJ, Chrome, Safari, Excel, Logic Pro among others all day long, heavily. Loads of other people seem to be having no problem with their systems. Eh.
If something has literally millions of users, saying "{X} happens" will invariably draw out someone with a similar anecdote, but that doesn't mean it happens to everyone, or even a significant minority. Some unique combination of factors is yielding a poor user experience, and hopefully there is a resolution, but it just doesn't seem likely that it's widespread.
elephant in the room for all these sort of articles it the law of large numbers. Sell a couple million devices with a defect that affects .1% of people and it'll be a thing for 1,000s of folk. And apple sells 10s of millions.
Unfortunately it’s comments and religious defenders like this that allow Apple to get away with selling broken products for years, and then issue recalls for products sold 4-5 years ago when people are not using them anymore, like they’ve done with multiple Mac laptops that had graphics card and screen issues.
Yep. I bought my M1 13" (with 8gb of ram) when it launched and it's easily the best laptop I've purchased. I want the 16" with the M1 Max because it looks incredible but so far the M1 13" has chewed through everything I've thrown at it, and I use it heavily every single day. Fantastic laptop.
Im willing to bet this community has a higher percentage of apple users than most others. The number of people claiming to already have an M1 of any kind is surprising, especially considering Ive received a macbook pro for every single software engineering job Ive ever had.
I think the main reason why these stories are brought up for Apple so much is because it has this persistent "things just work" narrative that a lot of people buy into when they decide to pay the Apple premium. When it turns out to not be true, it stings more than similar issues on e.g. Windows or Android, simply because the expectations there aren't that rosy to begin with.
Please don't make spurious generalizations about the community—these images are nearly all in the eye of the beholder, i.e. people with the opposite preferences to yours see the community exactly the opposite way (and make similarly spurious generalizations about HN favoring Apple or whatever $BigCo they don't personally care for).
Because such generalizations are just encodings of personal likes/dislikes in the form of general claims, they lead to lame discussion.
I have had really bad issues with bluetooth mice on my M1 macbook, the lag is so bad. Bluetooth is basically unusable for me so I had to switch to the logitech unifying receiver which has been flawless. This seems to be related to using airpods at the same time.
I love that Apple hasn’t been able to fix their Bluetooth bugs in forever. I always hope they’ll be fixed with the next OS version, maybe the next MacBook, maybe the next iPhone.
Nope. My AirPods have regular issues and my Magic Mouse regularly disconnects.
Pipewire for some reason insists on using GTK4 for its UI though, which hangs X cold if you use any of the popovers in it. I didn't even know it was possible to hang X like that.
What do you mean by a Pipewire UI? It's a background daemon (or three: pipewire, a session manager, and pipewire-pulse).
EDIT: Perhaps the parent means Helvum, which ldd says links to libgtk-4.so.1 => /usr/lib/libgtk-4.so.1. However, I can't figure out how to open a popover (https://python-gtk-3-tutorial.readthedocs.io/en/latest/popov...) in the app; all I can do is drag to create or break connections.
Try to set up https://github.com/jaakkopasanen/AutoEq impulse responses for headphones, and you _will_ inevitably hang your X. Daemon isn't terribly useful on its own if you can't set it up to do what you need. So anyone considering Pipewire as a "better" alternative to anything should also know, at a minimum, that it's difficult/impossible to uninstall and buggy AF.
To clarify things, PipeWire is an audio daemon, completely unaffiliated with https://github.com/jaakkopasanen/AutoEq which is a set of headphone impulse responses. It just so happens that the recommended way to apply these impulse responses on Linux is EasyEffects, which now depends on both PipeWire and GTK4. Nonetheless, PipeWire is not EasyEffects and not AutoEq, and "Pipewire for some reason insists on using GTK4 for its UI though" is untrue because EasyEffects is not PipeWire's official GUI, and doesn't even come up when I search `pipewire gui`.
I haven't tested EasyEffects. With that in mind, PipeWire is buggy AF for the time being, and I've been working over the past week to clean things up (fixed a few bugs in plasmashell and pipewire so far, but I'm still learning and there's a great deal of bugs I haven't diagnosed and explained to the maintainers, it will probably take weeks longer for me to work through everything, and I'd very much prefer the program was written to handle race conditions and edge cases properly, instead of leaving users to discover bugs and trace the root cause through the labyrinthine IPC-heavy event-driven codebase after the fact).
My guess is that Pipewire is written in C because the author likes C (or at least somewhat dislikes C++). And it has "abstraction"... in the form of hand-baked vtables and observers created out of nested macros, void*, and intrusive linked lists.
An advantage of C is that it comes with a stable ABI, where Rust is unstable and C++ is a subject of conflict (though you can achieve a stable ABI by restricting your external API to C functions, and if you want cross-shared-object virtual calls like Pipewire, you might have to write the same pile of macros as PipeWire, which you were trying to save yourself from). The downsides are memory and type unsafety (surprisingly the only crash I've seen so far is a null pointer dereference, and I haven't hit any obvious memory or type errors), and a complete absence of "vocabulary" or templated types (they've created their own slotmap-like ID allocator but without a generation count resulting in ever-present "logical use-after-frees" in library and application code, and an intrusive linked list which they use even when hashmaps might provide somewhat faster iteration and asymptotically faster search).
I assumed the C ABI, and thus the cobbled-up vtables. Where the author might have used C++ to advantage would be automating other, internal details of the code. There is nothing quite like bugs you don't have because you didn't need to write the code where they would have been, or because you could run code at compile time to generate or at least verify your cobbled-together apparatus.
The saving grace is that it can at any time be switched over to be built with a C++ compiler, and then start to be modernized incrementally.
What you say is true. Unfortunately I don't think you can trivially switch it to a C++ compiler, because they use {[FOO] = ...} initialization.
It would be cool (not necessarily useful, probably not a good use of time) to rewrite it in Rust while keeping the dependency count in the single digits. It would be helpful to contributors to add architectural documentation.
C++20 added dot initialization, which has been supported in MSVC/GCC/Clang for over a year now. No released version of the C++ standard supports array index initialization, and jcelerier says Clang but not GCC supports it as an extension.
I finally figured out how to get bluetooth headsets to work correctly on my desktop: output to 3.5mm, 3.5mm to dedicated bluetooth transmitter.
Throw hardware at the software problem, because software absolutely cannot be trusted not to foul this up. Not Windows, not linux, not pipewire, not pulse, none of it. Evidently not mac either, though I've personally never had trouble with bluetooth on mac. Bluetooth is 23 years old and its flagship application is still completely broken across huge swaths of the client/server compatibility matrix.
I switched after seeing PRs improving Pulseaudio get shot down pretty dismissively, and iirc the would-be contributor getting fed up, abandoning it because well whatever, Pipewire.
But actually, since switching the AV sync is off using my Bluetooth earphones, at least in Firefox.
Curiously, I can fix it with latency adjustment in pavucontrol, but it only has an affect when the 'advanced' drop down is revealed in that tab of the menu. Close it, and it's off again. Open, and the value has persisted, and it works again. Haven't really dug into that because I've been hoping the (presumably) underlying Pipewire issue will be fixed.
Oof, I saw that PR drama too (~half of bluetooth headsets were broken, PR fixed it, got ignored for months despite loads of 3rd party attention, author got heated, maintainers show up very interested in swinging their CoC around but very uninterested in merging the fix, author gave up and left for pipewire).
Pulse is still broken a year later, pipewire is fixed.
I think you're right about this. I have a pair of JBL headphones and it will just randomly become completely garbled during a video conference under Windows.
I have to disable/re-enable Bluetooth to "fix" it. It's a mess.
Conversely, I have a couple of Bose QC35, and either one of them works with pretty much any device I tried to pair them with - Windows laptops, Android phones and tablets, and a Chromebook. So perhaps it's a QoI issue?
I've had similar issues with Bose QC35 on Windows. Currently they work -- but the settings are on a knife-edge: all it takes is an unlucky Windows update and I'll spend half a day trying to find the right services and settings to reset/reinstall to get it paired again. (Also, whenever I restart, I have to open the bluetooth troubleshooter to get them paired, though at least that always works.) I blame both Windows and Bose equally, but I also think it's a very difficult protocol to work with.
I've had an issue with Windows 10 cutting off the first 0.3-0.8s of any audio that plays over Bluetooth for a while now. It's occurred across entirely different systems with entirely different bluetooth chipsets and entirely different headphones. Nothing is a surefire fix except for keeping a silent audio file playing in the background, which is ridiculous.
The bug isn't that bad if all you use bluetooth headphones for is YouTube, video chat, and music but it makes Anki language learning decks with audio almost unusable.
I think it's not that it isn't "good", it's more that it is an enormously overcomplicated design-by-committee. The spec is really complicated and downright weird in a lot of places. Kind of like they were trying out new academic ideas that weren't really practical.
I'm more familiar with BLE, which is better, but even that has a completely weird "characteristics" protocol that you have to use. Want to send a big blob of data? You have to manually cut it up into 20 byte chunks to send it over "notifications".
It's like they looked at the OSI model and thought "ah yes this looks right! we'll implement all the layers!"
Anything below Bluetooth 5 is barely usable. The 2.4Ghz freq gets interfered by anything from fluorescent lamps, old Wi-Fi hotspots, old laptops and phones to microwave ovens, and if it's not yours, it's somebody else's. You basically have no control over whether your bluetooth devices get interfered.
I used to have similar problems. It turned out to be Bluetooth congestion. When I relocated to an office away from a window that faced a public street, the problems went away.
Both in this article and other comments here, people mention Safari being a factor. It sounds like this is likely related to the browser. I noticed that you're using Chrome. I use a variety of browsers but rarely Safari, and I don't have any of these issues either.
I find all three (Linux, MacOS and Windows) to be fairly stable. It's been many many years since I've run into either BSOD on Windows or the beach ball of death on Mac.
My guess is faulty hardware. I've had plenty of those on the 2018 MBP (keyboard getting stuck, bulging battery and dead zones on touchbar). Waiting to see if the QC is better with new M1 Pros before buying.
>What I do get, is some bluetooth issues (e.g. affecting the mouse cursor. If you start the bluetooth deadmon it fixes itself for a day or so.
Can you elaborate on this? I think i was hitting this as well. Specifically when I scroll or type, it "halts" for a split second every other second. Originally I thought it was a video card failing or something but after a long time I finally traced it to Bluetooth. Since then I have been using wired connection on my keyboard and mouse.
What do you mean by "If you start the bluetooth deadmon"
In my case it's visible in the mouse (the keyboard is wired, mechanical), and it's like you describe, at certain points there is laginess. After some time, it's like the mouse slows down, and lags more.
I've heard this happens when you have some USB-C hubs attached, which I do. Not sure what the connection is -- but I think my Anker hub does causes this.
>What do you mean by "If you start the bluetooth deadmon"
Sorry, I meant: "if you restart the bluetooth daemon", e.g. "sudo pkill bluetoothd".
a little unclear if you don't have a hub, or do have an Anker hub, but try moving the Anker hub far away from your Macbook and see if you still have the bluetooth issue. You need a USB cable (depending on what's the hub is you might need a 3.1 Gen2) long enough.
For me it was a hardware proximity issue. Powered external devices interfering with the Mac's bluetooth.
Same here - I've been slamming my 13" M1 base model (8gb ram, 512 gb storage) and it's a beast of a machine. I throw a lot at it for work every day - multiple browsers open, back-end processes, the UI app, pgAdmin and postgres both running locally...it crushes everything I throw at it, with great battery life too.
Only issue I've seen with it is some bluetooth flakiness sometimes the way you mentioned. Other than that, it's been a rock solid device and the only reason I haven't upgraded to one of the new MBPs despite really wanting one...is because so far I haven't been able to justify it yet considering how rock solid the M1 has been!
i have an older macbook pro (only for another week until my new one arrives), and i found switching from iterm2 to kitty improved performance of my computer drastically. i think i have some other HD corruption issues going on, but kitty extended the lifespan of my computer such that i was able to wait for the new pros.
This whole thread is striking a nerve with me since these seem to be all of the issues i had been dealing with since getting a new 16 inch a few weeks back. Specifically "windowserver" and also these bluetooth issues.
See my other comment in this thread about how i solved "windowserver".
Bluetooth issue was a hw proximity issue for me. My CalDigit hub was mounted right below my Macbook. I noticed when I moved my MacBook's location (still connected to hub), the bluetooth issue when away.
It's worth a try if you have a powered accessory near your macbook and haven't tried this already, but...
I do still get the issue once in a while, but I can now at least just turn the keyboard/touchpad off/on and it goes back to normal whereas before, it happened all the time.
Once in a while I still have to Shift+Option+"bluetooth menu icon" and reset bluetooth module.
I know these are just work arounds, but at least it allows me to get back to work and only have to deal with it weekly instead of not being able to use bluetooth at all.
Edit: just upgraded to 12.0.1 and noticed Shift+Option+"bluetooth menu icon" to reset bluetooth module is gone!?
It’s probably USB 3.0 frequencies clashing with Bluetooth freqs. Even worse if you’re connected to 2.4Ghz wifi since it’s the same frequency Bluetooth uses.
If windows is not responding error it's an app bug and not a windows bug. I have to swap between windows and mac all the time and both are perfectly usable when you have enough memory for your given task, and that is 95% an app issue as too many modern apps drink memory like they just came out of the Sahara.
Explorer hosts a million plugins for everything that wants to do things like add a menu entry to the desktop right click menu, and each of them can crash or hang Explorer.
There’s similar interfaces for Finder for things like Dropbox, except not every mouse driver has a plug-in and if it turns out the model is problematic, like on Windows, Apple has no issues with repairing it and forcing the plugins to update.
Backwards compatibility is nice but it drags you down.
Since we all seem to be sharing that we don't have issues with our OSs, I've never had much issue with any NT based Windows systems. I've had bad hardware that once replaced improved reliability, but generally speaking I can go months without reboots. (except for those corporate forced updates for security reasons, but those exist on every platform, so I don't consider that an issue)
On the last few versions of Windows, it doesn't really feel like my computer anymore. I use my OS to lay out the structure of my thoughts, ie. I'll have text documents open, arranged in a certain way, project folders open, image editors with sketches in them... then Windows Update wakes my computer while I am sleeping and force-closes everything; I awake to a blank slate. It's as though someone has snuck into my house in in the middle of the night and swept everything off my desk and into the trash.
Anecdotally, it feels like I must restart WindowServer in OSX more often I've had to restart explorer/dwm in Windows. But neither are particularly reliable, in my experience.
Do you by any chance have the base model? I'm trying to see if folks think it's sufficient for their needs. It's hard to calibrate memory needs coming from Intel.
I have the base model 13" M1 (8gb ram, 512 gb storage) and it's rock solid and sufficient for my needs. I use it every day for dev work. Got my UI app running on it for local development, backend processes, connections to Heroku and my local postgres db, multiple browsers open, music streaming, multiple bluetooth devices connected to it...
Performs like a champ with excellent battery life. I keep waiting for it to start giving me some slowdown as our apps become more complex, so I can upgrade to one of the shiny new laptops...but so far, it's been a champ.
Are you worried about swap use at all? I’m on the same boat but I’m a noob so I just put my IDE and browser tabs on my memory. I don’t even know how to use docker yet
Not at all. I haven’t had issues with swap use so far and I intend to upgrade on a regular cadence (every 18-24 months) so I don’t expect it to be an issue for drive longevity.
It's probably worth reminding everyone who is new to the Mac or like most of us keep forgetting. Don't update your OS until at least half way through its update lifecycle e.g. for Monterey wait until 12.0.5.
This happens every single year where there are annoying/serious bugs that takes a while to get fixed as the focus is on P1 showstoppers.
Yup. I upgraded from High Sierra around the time Big Sur rolled out.
To be honest, I’m probably not going to upgrade again at all until the new MBPs seem stable enough for me to buy. Nothing in Big Sur or Monterey seems compelling enough, apart from M1 support and access to iOS apps, for me to risk the disruption.
> Still an option. I updated to 10.15 "Catalina" (released June 2019) about a year ago, and am still on it.
On new hardware released with Big Sur? To be clear, what I was saying is not an option is “downgrading” new hardware to an older major version than the one it was released with.
Hah well with the M1 air I got a few days ago, it helpfully won’t let me upgrade to Monterey! I’ve downloaded the 12GB update three times now and it fails each time it tries. I guess I’ll wait.
I do that as well - I have work Mac and personal one, and I don't want to run them with different OS versions since Apple loves to change things between versions for no reason (last time they renamed "grab" to "screenshot" and needlessly reordered System Preferences) so using two versions of Mac OS is super annoying. And corporate takes a while to get all their crap running on a new version.
I saw this recently primarily with Firefox and (presumably?) Firefox may have been updated to fix the issue because it is thankfully not happening anymore.
When it did happen though, it seemed ludicrous: that a single app could in a matter of literally minutes start burning through gigabytes of memory, to the point where the entire system seems to freeze with no option except to kill Firefox (and everything it may be doing) to continue. And, like the author, I had started keeping Activity Monitor open so I could pre-kill Firefox at a point when I knew it would probably start to be pushing it.
It never feels good to baby-sit a machine. It feels even worse when it’s an expensive new machine. And it means that the operating system is failing at its most basic task (arguably the entire point of having an operating system) — to manage resources well.
> And it means that the operating system is failing at its most basic task (arguably the entire point of having an operating system) — to manage resources well.
Before about 2001, that would be a bold claim indeed in the Mac world. Pre-emptive multitasking only became a thing with the transition to BSD-based OS X
Sounds very similar to the macOS 10.15.6 release, which had a horrible kernel memory leak. It was especially visible when running virtual machines. After a few hours the kernel memory pressure was so high that it would oom kill every usermode process one by one, including critical system processes.
That one had me yearning for a linux grub menu entry for "previous kernel".
It's fundamentally the same thing, but when the kernel is refuses to deallocate memory you can't just terminate the process, you have to reboot the system.
MacOS has had intermittent memory leaks for quite some time. Even my 2013 MBP running an old OS version has all the Apple processes slowly leaking memory until I reboot.
I doubt it's directly hardware related, though it's possible it is scoped to special MacOS logic that only is executed on M1. My assumption has been that it's actually coming from shared libraries that Apple-owned processes share, because I usually don't notice it happening for non-Apple processes.
Which OS version? Personally I'm on 2015 MBP with High Sierra 10.13.6 and zero memory leakage. I go months without even turning off the MBP. Granted, sometimes I get memory leakage behavior from Chrome/Firefox. I usually stick to Safari.
No problems for me with my M1 Mac. Heavy use of compilers, numerical tools (both native apple silicon and some that still need to run under Rosetta), a couple different editors, some specialized analysis tools, and the usual set of desktop apps for day to day work and fun. My workload tends to be pretty demanding of a machine (modeling+simulation where I can soak up every core and bit of RAM available to me). I can’t remember when I had to reboot the machine other than OS updates.
For people who are experiencing a leak on macOS 12: Check if you have set a custom cursor color in the system accessibility preferences. Resetting the cursor color to the default may fix it.
Having worked at Apple adjacent to the likely responsible teams, the relevant details are: overworked engineering teams, insufficient QA investment, and most of all a yearly macOS major release schedule that persists primarily to benefit the egos of Apple's upper management.
Until you see a WWDC that doesn't announce a major release of macOS, expect more of the same.
To any involved Apple folks here, I feel your pain, and I'm sorry.
And this is why I have completely stopped using all Apple products. I used to be a massive Mac fanboy. But the software quality is absolute shit now and has been for at least the past 5 years. I don't understand why anyone puts up with it.
I speak only for myself, but after Catalina I started transitioning to Linux and couldn't be happier. Being able to set up my entire computer with a single Github repo is a godsend for productivity, all it takes are a few keystrokes and I've got a fully configured desktop with all my dotfiles, keybinds and applications.
I've been a Linux user for some 10 years, but have also had to use a Macbook Pro for a while due to work. There's nothing stopping you from fully configuring macOS through a single repository similar to how you do Linux. For a while my dot files were aimed at Linux (primarily), macOS, and Windows as I anticipated to have to use both macOS and Windows down the line, though I've since removed much of the multiplatform code.
Things may be different now as I haven't used a mac since about 2015, but depending on what tools you needed you couldn't escape the GUI completely (for a fully automated setup). At this point I don't remember all that was wrong, but one of the biggest required installing xcode or something like that before I could get git to work.
You don't even need --cask anymore. You can also use `brew bundle dump` to synchronize what you've installed via brew, and it also can sync mac app store installs too.
It's a bit annoying getting macOS system preferences synced still although.
How are you managing the desktop (gnome) files? I swear every time I try to move them to a new system or account all I get is grief; weird bugs, application crashes, dbus weirdness. I would love to see your recipe :) . All I can usually do is bring over app config files like emacs/vim/etc/personal scripts.
I don't use GNOME on any of my systems anymore (the 40 update destroyed all hope I had for the desktop going forwards), but I'm convinced you can get some settings transferred by poking around in your ~/.local directory. Having written a few GTK applications in the past, it seems like the rationale for this is because of GNOME's settings API, and how they define compliance with the GNOME spec. I find it to be a total clusterfuck, and if I had any confidence in the desktop's current management team I'd probably open a few pull requests/RFCs to try and fix it. I might sound a little heavy-handed here, but the "my way or the highway" rhetoric that GNOME's developers are pushing right now makes it really hard for me to take their desktop seriously, or blame anything but the developers themselves for their lack of features.
In any case, I can assure you that behavior is perfectly "normal" to them, unfortunately.
That sounds really neat is it some thing you can share? I recently setup a new machine which didn't take very long, but I have run into a few things I forget so would like to set up something similar.
I basically just run an install script which does a few different things. I start by updating the repos and regenerating my mirrors so I get the fastest downloads possible. After that, I install shell utilities (my editor, shell of choice, base-devel, etc.) and then I enable the AUR so I can grab the rest of my desired apps. The rest of it is decidedly basic, it just copies my tracked config files to ~/.config/ and moves my wallpaper to /usr/share/wallpapers. I run a few rain dances automatically too, like `sudo chmod a+wr /opt/spotify` (which lets me automate my Spotify theming process) and installing/unzipping a Steam theme.
To install, I run `git clone https:github.com/username/repo && ./repo/install.sh`, and I'm off to the races. I really reccommend writing one for yourself, as it's a great way to learn shell scripts.
For me there's a trade-off, depending on how often I setup a new machine. Value of automating with a procedural script vs taking the opportunity to try a leaner (or more modern) toolchain.
Currently upgrade about every 2-3 years, so I do tend not to need a bunch of things each iteration, and might upgrade OS. But I'd like a backup in case e.g. hardware failure, laptop lost or stolen.
Some declarative workstation config like terraform or Ansible would be interesting.
I've been seeing this exact same comment for over a decade now.
The first couple of releases of OSX are always a little rough around the edges. And then by the end of the cycle everyone proclaims how wonderful this release is.
Except the bugs have continued to get worse and worse. For me the deal breaker was the blank page in iOS Safari[1]. After fighting that one constantly, I threw in the towel. I've been so much happier. Btw I was an Apple user from System 7 through to 2019.
Back in the days OS X was a dumpster fire in terms of overall reliability until the very late releases of every major version. Some versions like Lion had pure crashes until the end.
Even looking a few years before, we had the root user password bug and other pretty severe issues way worse than what we’re seeing now.
I did say over the past 5 years, so things like the root password issue I include in that. I'm also utterly convinced airdrop is just a long running joke at Apple, I've gotten it to work like twice.
Snow Leopard was as solid as OSX ever got in my estimation. Man I miss those days. Snow Leopard's primary goal was just to harden Leopard and not really have too many new features. Apple really needs to bring that kind of thinking back. I'm happy to see their hardware is back on track, now if they could just get the software to do a U turn...
What I don't get about apple, with a compensation policy that is less than the other big techs, because they can't get over the late 90s, team sizes that are usually smaller and being one of the largest market cap companies in the world, why don't they hire more?
Hiring more people isn't going to make them more money. Improving their products/services isn't going to make them more money. Providing more value to customers isn't going to make them more money.
Charging $300 for a screen replacement will make them more money. Selling bluetooth headphones with non-replaceable batteries will make them more money. Collecting taxes on developers will make them more money. etc etc
They got popular because of their commitment to quality. "It just works" is and in many ways still is a key to their popularity. They whittle away at their "quality credit rating" the less and less they invest in these things, until their quality credit rating becomes low enough for a competitor exploit that and make them lose market share very quickly. It does make them money.
I can see the business side of the permanent upgrade hype, constant new features, etc. But aren't we supposed to be reaching some kind of detente with Apple where they realize that power users drive the Mac market, and the release schedule should be built around hardware and stability, rather than endless retreads and new consumer features we'll never use? I'm still on El Capitan and pretty much everything since then looks like bloat to me.
You're right, I've just checked and when I set my cursor to be a custom color in accessibility settings suddenly my Firefox started to eat up memory as crazy :D. This went away as soon as I reset the cursor to default colors.
Haven’t been bitten by memory leak issues yet, but I feel for you. I had an M1 mini for nine months (lightly used) and had no problems. Switched from an i9 MBP 16” to an M1 Pro 14” a week or two ago. Have had zero problems and it’s the best computer I’ve ever had and fastest new computer I have ever used. MATLAB is 2 times fast than my Windows desktop at work, which was substantially faster than my i9 MBP. Oh and MATLAB isn’t even native yet, but running under Rosetta 2. I also have never heard the fans, even with two 4K displays attached, which ran all the time on my i9 MBP. These chips and machines are incredible.
I'm also experiencing this on 1 of 2 M1 MacBook Pro Max 32GB machines. I've performed a full wipe/restore on the faulty MacBook and the issue persists. WindowsServer is usually the faulty process consuming 20+ GB of RAM. The only difference between machines is one uses icloud and the other doesn't.
Given the similarities between both machines and application usage, I can't help but suspect some type of hardware issue is plaguing a percentage of machines out there.
You didn't tell us the most important part -- does the machine exhibiting the memory leak problem use iCloud or not?
Assuming the iCloud-enabled machine is the one with the problem, it seems odd to jump to the conclusion it's hardware? That's a lot of extra code running on the machine; not just iCloud itself but every single iCloud-aware app. There's a lot of additional potential surface area with iCloud enabled.
The iCloud enabled MacBook is the one leaking memory. I'm leaning towards a hardware issue because the iCloud setup is the OS default and the number of reported issues is still low.
Edit: A recent reddit thread may have correctly identified the WindowServer leak being related to the amount of full screen video played. My impacted MacBook is often used for Netflix/Hulu content. https://www.reddit.com/r/apple/comments/qos5n5/memory_leaks_...
"I don't understand how virtual memory systems work, but I know that if I write an article complaining about <new Apple thing> I'll get a ton of traffic".
The Activity Monitor screenshot in the embedded tweet even explains what's going on. The system has 64G of physical memory. A bit less than 50GB is in use (so 14G is completely idle, wasted). 7G are wired (kernel, device buffers, etc.) and can't be paged. 38G are in used caching things for applications, and about 180G worth of compressed data (mostly zeros I'd say, since it's only using 5G of real memory) has avoided being paged out.
There is literally no memory pressure in this system, and no evidence of leakage in those numbers. Darwin deliberately tries to keep as much DRAM as possible live to minimise page-in and decompression costs; it literally can't find anything worth paging in or decompressing to fill those last 14G. That's a vacuum, not pressure.
I'm not going to debate the stability issues that the article author is complaining about; I do hope they have all of the crash reporting features enabled so that their crash data can be used by the various developers to help fix things.
But I would hope that before writing yet another stupid clickbait article, they pick a topic they know something about.
The 38G isn't "cached app memory" (there is no such thing), it's actual resident memory, calculated as the sum of the second column. The same as "RES" in Linux. There is no memory cache that differentiates apps from non-apps (whatever that is?), there is just regular file memory cache, which is labelled below the application memory section, shown to be using 13G.
Every file you open on your system (read: app you open, file that macOS accesses, etc.) will be automatically read into the file cache and left there until some event, such as not having any unused memory since all of it is used as cache, requires some contents of the cache to be evicted to make room for an application.
If this system had been low on RAM, this "Control Center" app would have been paging to disk to make room for the gigantic 26G it "needs" (the fact that it needs this much is a bug/leak. It should not be showing a number that high in that field, which is the resident memory required by the app.)
In summary, unused memory is meant to be used for caching files (see https://www.linuxatemyram.com/). The cache is lowest priority and almost considered "free" memory, in the sense that the cache can be evicted anytime to make room for actual application resident memory. In this case, "Control Center" is using 26G of resident memory, not cache (which is not shown in that column at all.) This is a bug/memory leak.
No, first of all, there is a screenshot in the article showing high memory pressure. That needs to be taken into account by your analysis of the author's intentions and understanding.
Second, I have one of the new new systems with a lot of RAM and when WindowServer takes up 20-30GB of RAM, and other apps take up 2-10GB more than they should, I see high memory pressure. There are other reports of this. Don't extrapolate too much from one screenshot showing a system shortly after reboot.
It's possible that most of the bloat from this bug gets pushed to swap and harmlessly stays there until reboot, to leave space in DRAM for actually useful application memory, but the typical user doesn't have guarantees of this based on the instrumentation available to them.
The second screenshot literally shows safari occupying 30+ G of memory. What the hell is it caching? The entire internet?
Or maybe is it just memory leak?
for all of apples issues, i really respect their continued commitment to a personal assistant that doesn’t collect all of your data. as opposed to Am & Alph.
This is technically true, but only in a pointlessly narrow sense. They do not sell your data; they hoard it, and sell ad targeting services that use it. They also provide it to the government, for free, as required by law.
But does anybody actually care about this distinction?
it’s tiresome to you because you are a liar fighting a tide of truth. if google pays your bills i can’t say that you should definitely be ashamed of yourself, but otherwise you should definitely be ashamed of yourself for trying to mislead people and be as duplicitous as possible at the top of your lungs. i find cultists really tiring.
It’s significantly better for battery life, it’s a tiny bit more integrated into the OS, and as a bonus it doesn’t send every aspect of my entire browsing history back to Google.
Battery life, better responsiveness, less google spying. If you're not an extension power user and don't mind it's tab overflow behavior, it's pretty good.
That's an interesting point. I notice that the person who wrote the blog post is running Safari, too. Is this a Safari problem? The activity monitor screenshot makes it looks like Safari is eating up big chunks of memory for each tab. I don't often use Safari, and my M1 MBP has been problem-free.
I have no proof but a strong gut feeling the root cause of this is broken OpenGL drivers. Here's why:
Over the past several months we've been very busy porting our Android- and Windows-based 3D Capturing app to iOS. We have quite a bit of OpenGL (ES) code that needed to be migrated. After a couple of initial hiccups things appeared to work well, however something was off. OpenGL code that's fast and completely unproblematic on Win/Android caused weird CPU stalls and after some minutes of app usage typically everything just crashed. After some digging and debugging it became clear that certain OpenGL function calls were leaking memory like crazy, as in several MB per call. Concretely I'm talking about glBlitFramebuffer() and glCopyTex(Sub)Image2D() . Both of these operate on frame buffers (memory that's often as large as your app window size) and they were internally allocating new buffers on every call and apparently never freed them.
Now of course this is iOS and the article is talking about MacOS but it's quite likely that Apple derived their M1-targeted OpenGL implementation for MacOS from the iOS version which has always targeted Apple SoCs.
Given the amount of OpenGL-based software still out there and Apple's deprecation of OpenGL in favour of Metal (they just don't seem to care any more) those kinds of large scale memory leaks don't really surprise me. OpenGL is still pretty much everywhere and if you have bad memory leaks in core framebuffer operations things start to accumulate very quickly.
FWIW, I notice the memory leaks particularly with “Control Center”, which is an Apple extension to the task bar. I wouldn’t expect any Apple-written code to be using any OpenGL; it should all be in Metal.
Yes, I see that now. Thanks. There seems to be a lot of confusion about what parts of Carbon are and are not going away or did not make the 64 bit transition. I haven't tried to build a Carbon app in a while. When you look in the frameworks, the ones that are present now seem to have arm64 bridge support, which I assume means arm native code for running Intel apps in Rosetta. Perhaps some very large user base app vendors like Microsoft and Adobe have not quite let go. Probably some old Apple code also. Apple keeps announcing it is disappearing but it can't seem to be killed.
No offense, but I laid out exactly what the framework is used for in my original comment: https://news.ycombinator.com/item?id=29144758. It's not (exclusively) there for Rosetta or third-party compatibility.
If anyone wants to do any fun OS X GPU rabbitholing, this command will dump out a bunch of stuff from the IO Registry as an XML/plist:
ioreg -c IOGraphicsAccelerator2 -r -l -a
The PerformanceStatistics and IOReportChannels dicts are particularly juicy, and under IORegistryEntryChildren you can see which processes (IOUserClientCreator) are using what OS-level GPU resources, including Metal vs OpenGL API.
It's unclear to me if these issues are happening when you are running the iOS app on your Mac (emulating during development)?
If so, try setting Mac to use the discrete GPU all the time and see if it keeps happening. MacOS flip/flopping between discrete and integrated GPU was an issue for me. I can only imagine what MacOS might be doing during development of a graphics app such as yours.
I've been dealing with this ever since a software update to the M1 MBP last summer and it's been an absolute nightmare. It's obviously some kind of system-level memory leak, because my workflow is just basic apps (Superhuman, Chrome, Slack, Spotify, and WhatsApp are the only apps running most of the time; with the occasional other stuff like Keynote).
I used to do software development, running a million things at once and in decades I can't remember last time I saw the out of memory error. Sure, slow downs due to swap, but never straight up the error forcing me to kill apps.
On the M1 MBP, this error modal would pop up dozens of times a day. I would literally be constantly be quitting any app i'm not directly working on. It'd take a few days to get to this point, but at this point nothing other than a reboot would get it back to normal. The memory total for the apps running (plus the background stuff like Dropbox) was way below the built in 16GB, not to mention zero swap to SSD. Somehow there must be a leak in the OS.
Currently sitting at:
WhatsApp Helper (Renderer) 3.57GB
WhatsApp 1.37GB
Dropbox 1.13 GB
WindowServer 1.05GB
etc.
My main pain point aside from random OS crashes is Bluetooth / audio.
The sound device will just drop working randomly whether 3.5mm Jack or Bluetooth. Restarting the service sometimes works sometimes not, at first I thought it was MS Teams but didn't seem app specific.
Makes video calls painful knowing either sound or mic could just stop working at any time.
Bluetooth issues are often associated with endpoint security software such as jamf, crowdstrike and carbon black. If these issues are happening on your work laptop especially, you might have those things installed on them without you realizing it.
Apple's attitude towards endpoint security software is basically, "it's stupid and you shouldn't have them installed, our recommendation is to uninstall them" so that's that.
"Third party “endpoint security” software may cause slow simulators, system freezes, or prevent debug processes from running in simulators reliably. This sometimes manifests as debugserver disconnections or simulator applications receiving a SIGKILL signal. (55853555)
Workaround: Uninstall the third party software."
Note that this also was when Apple released the EndpointSecurity framework specifically made to sandbox these sorts of corporate malware in user space and outside the kernel.
I've had my 14" M1 Pro for just over a week now. It has performed amazingly. But I also don't install tons of random apps and let them run in the background. I never restore from backups, I manually transfer files from my user folder over to my new setup. I always check whether an app has native M1 support before downloading, though I understand this isn't an option for everyones workflow yet. In cases where the app is literally a wrapper for a website like discord, I just use the website. Or in the case of zoom where I just don't trust the app to be on my computer and the website is sufficient. If something I used previous doesn't have native support I look for an alternative and a couple times I have found something better and/or open-source. I understand rosetta 2 has some issues with overusing swap, so I would rather avoid degrading my ssd and experience.
So far this has worked for me, and I'm able to get two full day's of battery life. I'm away from my desk far more frequently now.
Edit: My control center is using around 1GB of memory. Low enough that it wasn't causing issues but still concerning. It appears there is a memory leak somewhere, hoping a bug-fix is pushed out soon.
My M1 Mini crashes about once a week. I look through the crash log before sending it off to Apple, but is always seems to be something different that takes down the system. In comparison, my 2018 Macbook Pro rarely crashes and is running the same version of the OS.
I have the M1 Pro Max on my to-buy list, but I'm going to wait a bit to make the purchase since I really can't have my $4000 workstation crashing on me.
I'm kind of flipped around. My 16 inch 2019 MBP panics every two or three days. My 13 inch M1 MBP is rock solid and memory pressure is fine. Both running Big Sur. I won't call the M1 machine my favorite Mac of all time (my old 2013 takes that prize) but it's solidly in 2nd place right now. It's given me so little trouble...
My M1 Mini crashes (while sleeping) about once a week here as well. Funny thing is that I never get the crash reporter showing up after a hard reset. Very strange.
Two or three times, my M1 MBP has crashed overnight, but with the same error - something involving external drives. I keep 6-7 external drives connected at all times via two chained hubs. Hard to know if it's the OS or a flakey drive.
Lightroom Classic (latest version) is unpleasant on my wife's M1 and even crashes. Last night I moved it over to my gaming PC which is pretty old, but performs much better.
It ran on her 2012 Intel MBP and whatever dayjob MBP I had in the past.
First, thank you for your reply. I am curious, after this experience, would your next machine be also an Apple with M1 (or the next iteration of it), assuming nothing else goes wrong?
It could also be that x86 has better SIMD support by a considerable margin, which can make repetitive memory access/serialized workloads a lot faster. I'm not super well intimated with those extensions, but I know that the NEON SIMD implementation in ARMv8 leaves quite a bit to be desired. It's a tricky situation, and one I don't see resolving in a nice clean way. It's stuff like this that makes me hopeful for RISC-V though, where we could theoretically have our cake and eat it too, with dynamic instruction pipelines and incredibly low power usage. Only time will tell, I suppose.
Throughput is not an issue on the M1, with 4x 128-bit SIMD units.
Neon is certainly not a bad SIMD ISA, it's a quite orthogonal one.
You also have the AMX extension at hand, which is more special purpose but allow to deliver very high throughput. (on a regular M1: 350Gflops DGEMM, 1.2Tflops SGEMM, without leveraging anything other than the CPU)
LR CC runs much much better on my M1 MBA than my prior 2017 MBP. I wonder if Adobe is just dragging their feet on updating classic since they want everyone moved to CC?
Sometimes it just crashes. This has rarely happened on Linux. This sometimes happen on MacOS to the point where I don't trust it. At all. Ever. I have to save everything I'm working on constantly, just in case.
Sometimes apps just stop working. Right now telegram just won't start. No matter what. Never happens on Linux.
Sometimes it won't wake up from sleep. Never on Linux.
Sometimes it loses the keyboard and/or mouse. This never happens on Linux.
Sometimes the screensaver works, sometimes it doesn't.
Grep is slow, waaay slower than on Linux. So are other random CLI things. How? Why?
Docker... don't get me started, and don't make excuses. (~700 people at Mirantis and ~150k at Apple can't figure this out?)
Everything runs slower on MacOS. Every. Thing. grep, sed, awk.
Why don't sed, awk, grep, etc.. all the Linux things work quite right? I need to add something else on macOS to get them to work like they did on Linux. It's always "almost" but never do they work like I'd expect.
I need to remember to turn off my monitor at night, because some nights it just randomly turns on. Never on Linux.
Maybe it'll wake up, maybe not. I figured out if I unplug my headphones and/or webcam I can force it.
All of these super basic things just DON'T work. It sucks. I could forgive Linux, but not an OS and Hardware made by the biggest company on the damn planet.
I get it, this is faster than a 4 year old laptop. But that's not what I'm comparing it to. I'm comparing it to a 5 year old system I built myself from random parts that runs an open system operating system. That's a pretty low bar, and it doesn't win.
So why use OSX? I've been running Linux for over a decade full time. I remember using osx at work and feeling like it was just clunkier, slower, and buggier.
I've gone back and forth between linux and macOS the past 7 years at work. Both get the job done, but there is absolutely no denying (in my experience) that macOS is slower.
Clunkier and buggies depend entirely on which distro you are comparing it with, though I find clunkier at least to be true.
Buggier than linux WMs and desktops is a pretty high bar to beat, as long as you're not including homebrew in the equation (worst, most necessary bit of software for Mac ever).
OSX is great, I love my M1 laptop. I think Linux is better but if you're on the go and not -too- into Linux that it is a fine substitute. All the Linux only coding that I need to do I can do in a terminal with remote login. At home I use Linux but on the go I have a work laptop with OSX and it's a good substitute, especially if you don't have to buy the laptop yourself.
Because last time I tried, I spent days to have fractional scaling on 4k monitors work without screen tearing, blurring etc. And have consistent momentum scrolling on touchbar.
I do not even think about these things on macOS or Windows.
At this point my time is too valuable for that shit. I used to fiddle a lot with my computers when I was 16. Now I want them to work.
Weird, I got an M1 Air recently and it's super fast (the most notably faster seeming new computer I've gotten in a decade) and all these issues are incredibly foreign to me. I just can't relate to this comment at all...
What use case do you have for your Mac? The benchmarks and battery life have been tempting me, but I use Docker a lot which I'm hearing mixed messages for, and to be honest, my Intel-based work Mac is less palatable to use than my personal Ubuntu setup, especially in regards to workspaces, and some weird Bluetooth issues.
I’m on the same boat. Using an intel Mac provided by work. Had an M1 MacBook Pro until a few months ago, and it was great at several things(fast, responsive, quiet, and great battery life) at the cost of crippled Docker use(I needed local docker at that time).
I feel like many many engineers like me would easily get a reliable laptop with MacBook Pro like specs. The problem is that simply does not exist.
Excellent hidpi display.
Ryzen 5000 series CPU with its low power draw.
Good build quality.
Reliable support system.
Non weird components such that a recent release of popular Linux distribution would “just work” without needing manual tweaks.
In my experience, you get 3/5 in the above list at best. Framework laptop hits several of those points but misses out on Ryzen. Another contender on my list is ASUS 14 OLED with Ryzen - I’m worried about their support system though.
If you can compromise on the HiDPI Screen, checkout the HP EliteBook 8X5 Series, they are amazing machines. You can plug up to 64gb of RAM into them if you need it, they run pretty smooth and my colleagues can run Linux on them without any issues.
Wow! Yes, this is exactly what I was looking for. Looks very appealing. I might get one! They're expensive, though. But it might honestly be worth it to me.
I want just to add that I am using a 8x0 g5 elitebook currently (because at the time there was not the 8x5 series which uses amd ryzens) and it's by far the best linux laptop I ever used in my whole life. Also very sturdy and looks great.
Well you know, it's like the standard of the high end for windows laptop, still behind what apple put in their macbooks, but comparable with what you are gonna find on xps and thinkpad x/t.
Looks nice indeed. I have been spoiled by the excellent Retina display on Macs. For something literally in front of my eyes for a big part of the day, it must be good. I cannot see myself compromising to a lower res display :-(
May be for others that are not as nitpicky as me….
Probably not any better either. If you want to run docker on a mac you probably should not, but the second best strategy is probably to bite the bullet and run an actual VM in which you do your docker stuff.
Odds are a proper and well maintained hypervisor you keep running will be a lot better than whatever DfM does.
Fair enough. My Intel Mac gets about a 5x performance penalty doing Webpack builds in Docker versus natively. My coworkers using M1 Macs report similar numbers. What workflows are you seeing M1 Macs getting 6x worse Docker performance than Intel Macs?
Eh, it is worse. Worse to the point that even though I just got a MaxBook, I won’t even conceive of getting rid of my 2020 maxed out Intel iMac. And it’s why I have other x86 machines too for Linux or whatever, because containers on macOS are rough but containers and Docker specifically on the M1 is really rough. It is clearly an ecosystem/use case that Apple doesn’t prioritize.
It makes the case for stuff like GiHub Codespaces or sshing into a remote machine via VS Code a lot more manageable. For everything else, the M1 is a beast, and even emulated, you can get a really good Linux experience (Windows too for that matter — games, obviously not withstanding), but containers? Nah. It’s worse than the already sad state on Intel Macs.
Yeah I'd carefully look at what your expectations are with regard to compatibility. Docker itself seems to run just fine but running X86 (AMD64) containers is pretty slow compared to normal macOS docker which is already slow.
I've loved my M1 MBA but I purchased it specifically knowing I have a primary x86 workstation to fallback on for heavier tasks and I'll often give up trying to install some dep/run some program on the MBA and switch to that. It still works great as a general purpose device and I do some dev on it but if you're looking to do all your dev on it, I'd be cautious.
Def got slightly burned by docker on M1 Pro. Inherited a python project with 250 deps. Building docker image (part of is compiling spacy) used to be ~10 minutes, now it's a whopping 65 minutes.
Apple refuses to ship gpl3 software for seemingly no sane reason (judging by the fact other commercial software vendors including e.g. windows don't have this issue) probably explains the command line issues; apples' tools are an entirely different codebase from the tools in linux distros. And unless you actively compare systems, you might never notice the apple versions are different.
It's been this way since OSX day 1 IIRC, so it's a little surprising the apple versions are still apparently inferior (if the grandparent poster is anything to go by, haven't used em in years), but it's conceivable the CLI just isn't an apple priority.
I wish they'd just give up already and copy as much (of the CLI-relevant stuff) as possible from linux, and accept that they'll need to give up a tiny amount of control over a small sliver of their OS - a sliver that they never wrote in the first place and also can't really change the api of, because that would break all kinds of things, which renders their ability to close source somewhat moot anyhow. And most users never use this stuff! The linux-conventional gnu variants are well maintained and usable in both MS world and linux world - having CLI scripts that are more easily portable across three OS's would be great. Apple's toxic control-freak nature really hurts here, and mostly just their users. I really don't get it.
Anyhow, there are workarounds (but e.g. homebrew is hit and miss), it's mostly just an annoyance they're even necessary.
It’s a perfectly reasonable reason: they were forced to open source their objective-C compiler that was built on gcc and decided they didn’t ever want to be in that position again. The BSD tools are annoying, if you’re used to the Linux versions, but they’re perfectly reasonable implementations of the relevant POSIX utilities (better than busybox, for example).
It didn't help that rms was using it as a "hah hah that time I put one over on Steve Jobs" a decade later when I saw him speak. No idea if he still does.
I sure hope apple has a better reason for shipping unnecessarily impractical software than some pet peeve by a now-deceased CEO. That would be kind of insane.
> They're basically achieving their dream now with Swift, though. I don't expect them to ever release that source. I don't think there is any precedent for Apple releasing source code after the fact.
I wonder, to what degree is this still the case today? Has Apple's position on Open Source changed in the past half+ decade?
I don’t think it’s changed about GPL: GPL3 made things even less friendly to what they want, so they’ve started doing things like switching to zsh instead of bash. LLVM/Swift are open-sourced, but not GPL
And here we should note that LLVM (by its creator who already worked on it at Apple at this point) was offered to GNU and the FSF in 2005 but were ignored/chided [1] until a decade later when GCC/GNU/FSF realized what they had been offered ten years earlier and then had to grapple with what they had rejected, once LLVM had taken hold [2, 3], including proposals to try to refuse LLVM patches to emacs, out of some sense of petty NIH thing [the whole thread 2 and 3 are part of].
Reading those threads makes those decisions seem pretty reasonable to me; this doesn't look like a petty vendetta at all. And while chris sounded optimisitic about including LLVM, he didn't offer to; he offered the glue code to allow the gcc frontend to talk to the separate LLVM backend - and doing so implied a raising a number of fundamental requirements, such as the C vs. C++ bootstrapping issue.
Accepting major changes like that isn't trivial; and a certain amount of NIH is healthy here. I mean, that's kind of apples motto too right? This was a technical patch set with unclear upside (if any), and potentially relevant cost - why would they accept that?
No, GPL2 made them do that, and even so they were ok with it and used to pay people to work on GCC for the Mac in the old days. They rejected GPL3 specifically.
I think the above phrasing is fair, and technically correct. You can closed-source a fork of GPL software. It's extremely common for companies do this while software is in development. It's the moment that someone distributes the software that the license requires source distribution. Some people/companies never distribute their forks. This is why newer revisions like AGPL were written.
Probably because of the 'provide the keys' and not sue for any anti-circumvention tech in the GPL v3. The anti-tivo provision. Not sure if it applies to standalone system utilities, but I bet it is an internal blanket ban on the GPL v3.
Apple is very fond of locking down their hardware. Google took a very strong stance against the AGPL-even banning it from their own public code repo service before they shut it down.
My M1 Air crashes everytime I wake it from sleep and then immediately watch a Youtube video in Firefox. I can reliably reproduce this. If I open the laptop so it wakes up, login, and open a Youtube video it will hard crash close to 100% of the time.
Firefox just changed how their full screen video playing works in the most recent released (version 94). Could be that. Try an earlier version of Firefox.
M1 Air here that I've had for almost a year with zero issues, and I also use Firefox and watch a lot of youtube.
I can't speak to m1 in particular, but BSD grep is known to be very slow regardless of platform. You can just brew install grep to get a fast one, but you'll still always have to always deal with avoiding the slow system grep.
Ripgrep changed how I search my codebases, it’s so damned fast I love it. I almost always use it when I’m trying to find a pattern or a bit of code I vaguely remember but have forgotten where it might be in a large codebase. Still returns almost instantly.
Some mac users seem to be cursed with bad hardware.
It seems to be black and white: users either have no problems, or ALL the problems.
I started to get "the curse" after 8 years with my 2012 MBP. It started randomly turning off. Not crashing, just turning off! I replaced the battery because that was the only warning I was getting. But it can no longer run any macOS. Even mavericks. It crashes randomly. It went from rock solid to fragile as an elderly lady with osteoporosis rollerskating on a cobblestone path.
Now it is my primary Ubuntu machine since the touchpad works so much batter than Ubuntu running on my Lenovo Yoga and it has better performance. Zero turning off problems. It's been on for months running a VNC server.
Yeah, it is very strange. I have an early 2014? MacBook Pro and it still runs today. 0 issues. That being said, it's now slow as hell and likely because of the battery. But the development experience on MacBooks is delightful.
Don't know about others but my M1 Macs are the most stable and performant Apple machines I had in a long time. My previous 16" i9 MBP was a disappointment (laggy/sluggish, running hot) compared to my M1 Mini and MacBook Air.
I have all the issues the parent post is commenting about. I thought I was crazy... but same happens to me. Simple stuff... why doesn't sleep to wake work correctly? Why are my peripherals schizophrenic? My Windows 98 Compaq didn't have these issues... and that was two decades ago.
> Why don't sed, awk, grep, etc.. all the Linux things work quite right? I need to add something else on macOS to get them to work like they did on Linux.
If you mean how they work, like command line arguments and such (and not just speed), that's because macOS is based on BSD and thus using the BSD and not GNU versions. That's not really a "not right" thing, it's just different, and would be the same if you used any other BSD like OpenBSD, FreeBSD, etc.
If you install GNU coreutils from Nix (and maybe some other package managers, idk) you can get them with their normal PATHs, at the cost of some possible incompatibility with macOS-specific scripts that assume the bundled versions.
I used GNU ones without the `g` prefix for a year or two and never ran into trouble because of it.
no, it's definitely "not right" -- mac os ships with ancient, crippled versions of the bsd utilities, afaict simply because it's bad for open source and because they can. E.g. `man sed` then check the date at the bottom
I'll bite and talk about centos, the 'stable' distro.
Take the example of 'tar' that slowly evolved and some time even changed its default behaviour. e.g. what to do when untaring a folder but there's already a symlink with the same name as the folder, do you erase the symlink and create a folder or do you untar the folder in the symlink? Between centos 5 and 7 this changed... Causing me some amount of pain.
hostname stopped supporting '_' in host names. Could have raised a warning, have added a '-pedantic' option but no.
hostname stopped assigning the permanent host name with the 'hostname myhostname' command and now you need to use hostnamectl. Could have left the 'old' feature and make hostnamectl whine instead, but no.
ifconfig the deprecated tool subtly changed the order of fields too and now the MAC address is not in the same place as before, breaking scripts that trusted that not to change.
df changed its syntax, adding or removing a new-line some time ago.
Yeah thinking back we shouldn't have used the standard tools and expected them not to change.
So, now, whenever I see in code review some bash person using 'standard' tools to get info that could be gathered through procfs or sysfs, I instead ask they use procfs, sysfs or ioctls and be done with it.
To be fair, the last two are more from a bad design choice by *nix in general to base everything on passing around unstructured text and hoping each tool's ad-hoc schema won't change.
That 'hope' was true for some time, I guess. Those systems had no problems with previous migrations from early fedoras to centos 5.9... When your last line of compatibility is procfs and sysfs and ioctls or posix calls (statvfs for free disk space), well I know what Linux devs think of their installed userbase's time. Yeah yeah free stuff for free alright. Still, if you'd paid for rhel, same shit. Thank your dogs we provide an /etc/redhat-release, heh.
And if course even Linux procfs is not without its surprises. Adding columns or lines was foreseen but some changes are more fun!
Take the evolution in /proc/meminfo that changed the meaning of some of the lines there over time (WTF?), whilst adding others (which was fine, expected, etc.). Of course I want now to include IO cache (that will be reclaimed if I ever needed to malloc) in my system's free memory indicator! Heh you can still get back the old value with this formula, what are whining about?
The Telegram issue is not specific to M1. I think it's a bug in Telegram, as it simply refuses to start on my Intel based MBP at the moment. Started this evening.
Funny, I've been feeling the same way about my M1 MBA. It's hard to deny that the newer silicon doesn't have my older Linux rig beat for latency, but when it comes to raw performance I was pretty damn disappointed. At risk of offending some people, it was exactly what I feared the M1 would be: a phone CPU shoehorned in to a clamshell body. There are advantages and disadvantages to running a machine like this, but overall I don't really find the development experience to be that much more compelling than a Raspberry Pi.
That being said, Apple still knows their audience. I imagine there will be a lot of chronic Twitter browsers and Pages advocates who fall in love with this machine, and I think that's fine. I do have to wonder what constitutes a "premium" device if not for the hardware though. Everything I do on modern MacOS feels like it's prodding me to go spend $8 on a tool that I'd have for free on Linux or even Windows. Booting up Big Sur makes me long for my old a1502 running Mojave and handling some frankly ridiculous workloads while staying rock-solid in terms of stability. If anything, I applaud Apple for coming straight out and telling me that the future of their devices isn't for me.
> That being said, Apple still knows their audience. I imagine there will be a lot of chronic Twitter browsers and Pages advocates who fall in love with this machine, and I think that's fine.
That is delightfully insulting. Real developers would never use Macs, of course.
If your M1 experience isn't better than a Raspberry Pi (the only issue I have on RPI is speed and memory limits) then something was wrong with your laptop. There is really no comparison, the M1 will destroy it in any performance demonstration.
But when it comes to raw performance I was pretty damn disappointed.
I'm curious what you mean by "raw performance"? I'm no Apple enthusiast, but AFAIK the M1 is the fastest laptop CPU on the market. There are faster desktop CPUs, but I don't think that makes it "a phone CPU shoehorned into a clamshell body" — it beats pretty much any laptop CPU in existence in the benchmarks I've seen.
> Everything I do on modern MacOS feels like it's prodding me to go spend $8 on a tool that I'd have for free on Linux or even Windows.
Yeah, that’s so annoying. Wanna browse a .zip file (not extract it)? Either `brew install file-roller` with its non-macOS UI, or use shady paid shovelware. Wanna record your desktop audio as part of a screencast? It’s so complicated [0] and often involves shady stuff [1] or software that costs $99 [2]. What about Windows? There’s (mediocre) ZIP browsing support in Explorer since Me, 7-Zip is free and trustworthy, and OBS just has a desktop audio input built-in (and I also could do that easily in 2005). Linux? You’ll probably get a working archive tool built-in (file-roller, ark, whatever) and OBS desktop audio just works.
> Wanna record your desktop audio as part of a screencast?
It doesn't have to be a screencast, it's kind of a convoluted process regardless of which operating system you're using [1]. If I'm not mistaken OBS is using the built-in virtual loopback devices available through pulseaudio, but MacOS does not include a virtual loopback device by default, hence requiring a third party one such as Blackhole or loopback.
> ...and often involves shady stuff
...what shady stuff are we talking about here? Blackhole [2] is open source software. It's about as shady as any of the (several) linux audio systems.
Memory management on iPad seemed to be terrible when I used one for a while. An app would start acting up, so you would take a screenshot or video and send to Apple's bug reporting tool right? Every time you switched from the bug reporting tool to any other app (to crop the images of the bug, etc.) it would run out of memory and kill the bug reporting tool, which wouldn't save any of the text you had entered in the forms, deleting all of your work on the bug report.
Also, every browser tab with text entered was a random crap shoot if you were going to lose it by going to another tab to copy something and return to paste. Nothing was predictable, youor work was just constantly lost of the device. Almost no apps seemed to persist things when the OS forced them to shut down, except maybe the highest rated dedicated artist and note taking stuff. But even Apple's own apps (like the bug reporting tool) often didn't bother.
I'm strongly prefer Linux (been using it last 10 years), but I spent last year on macOS in waiting for Asahi to become usable. I still curious about your experience.
Did you actually tried to replace Apple garbage version of command line utils with latest one from GNU? It's kind a obvioius that OOTB version of utils will be much slower since Apple stopped updating GNU versions once license been changed to GPLv3. And BSD version of utils they ship with newer macOS is also usually inferior.
I don't think this is actually super obvious to someone that isn't familiar with macOS and is used to using linux. I do think that if someone finds something peculiar enough to notice and complain about it that they should at least do their due diligence to ascertain why, rather than simply being upset about it.
As much as I would like to get one of those M1 Macs, this is exactly the reason why I will wait for the 2nd generation of their processors. My Intel MacBook runs rock solid and stable. Until the ARM Macs haven't reached that level of stability, I am not willing to make the switch.
This might be the case for some users and basic use cases like surfing the web and text writing. But there are still a lot of issues for pro users and developers. If you rely on such a machine for your daily work, its not mature enough.
Also with various package installer you can sub in all your favorite linux versions of things. I have l.env file that aliases them all to linux equivalent commands.
Because the "linux" versions tend to have more features and better help output? I've never seen anyone install Linux and be like "damn, I wish my coreutils were more obtuse" and install a bunch of non-GNU stuff. Like, I get that Linux is just another unix and the GNU userland isn't somehow more legit than others, but it's not like people are intentionally getting inconvenienced just to be annoying.
That's a developer-perspective issue. You're talking to a user here, and from a user perspective, the GNU utilities are generally strictly superior.
Take the example you provided, of yes(1). Most of the extra code relates to optimizations, and ensuring that the entire argv is echoed back rather than just the first argument. Because of this, even a very simple example like:
"yes hello world"
will behave differently on BSD vs GNU, with the GNU version likely being what users expect.
Because the functionality in this case is a strict superset, regardless of what you're accustomed to.
I didn't grow up on GNU either, but the GNU yes does everything I expect having used other unix systems.
> This isn't something you can quantify and say 'strictly superior' it's all just taste and opinions.
For a complicated tool, sure. But this tool is very simple. The GNU yes(1) accepts 100% of inputs that work on the BSD tool. It does a few additional things too.
Hey, look. As a systems guy myself I understand the aesthetic around clean and simple software. That's a totally reasonable preference.
Here's the thing: GNU userland isn't some poorly designed aberration. It's by far the most popular unix userland in the world and is very commonly ported to systems which may do things differently. Almost every other unix has a distro to supply GNU userland tools on it -- no other unix userland can make the same claim.
Not everyone wants it, sure. Especially people with perspectives broader than the typical user. Of course. But it's the most popular system in the world -- not some weird never-used prototype.
And in the case of yes(1), I think you would have to try very hard to find a reason for a user of the tool to have a negative preference.
If you really want to get technical, Linux is just a kernel. There are certainly implementations of it that are POSIX-compliant, but since most people use it with GNU systems it's a bit of a moot point.
> Docker... don't get me started, and don't make excuses. (~700 people at Mirantis and ~150k at Apple can't figure this out?)
The whole reason Docker exists makes complaining about its performance on non-linux operating systems asinine.
> Why don't sed, awk, grep, etc.. all the Linux things work quite right?
Because MacOS is BSD-based and those tools aren't "Linux" things. You can install the GNU equivalents easily if you dislike them.
> I could forgive Linux, but not an OS and Hardware made by the biggest company on the damn planet.
Software development doesn't particularly care about the fact that Apple is the biggest company on the planet. The pooled resources going into development of linux is quite substantial and yet it too still has plenty of software issues. LTS kernels can have hundreds of patches because software development is hard no matter how many eyes you have looking at it.
Can you distill that into the single whole reason you're referring to? It's unclear to me just reading that marketing page.
If it's "to provide a standardized unit of software that allows developers to isolate their app from its environment, solving the 'it works on my machine' headache", I don't see how it makes expecting good performance asinine.
Linux's ABI is stable, MacOS can definitely implement it... there's nothing in Docker/Linux containers in general, forcing MacOS to incur virtualization overhead.
> When running on Linux, Docker uses the resource isolation features of the Linux kernel (such as cgroups and kernel namespaces) and a union-capable file system (such as OverlayFS)to allow containers to run within a single Linux instance, avoiding the overhead of starting and maintaining virtual machines.[12] Docker on macOS uses a Linux virtual machine to run the containers.
And MacOS is built on Darwin; a hodgepodge of FreeBSD atop Mach and years of iterating on that.
Point being, Apple has zero barriers to making Mach look like an existing interface to userspace a la FreeBSD. They can take advantage of Linux's stable ABI and add the necessary Linux interfaces for native Linux containers support, and it could be performant.
The whole x86 on ARM problem seems like the more difficult one to make performant from where I'm sitting, but there's no shortage of claims to the contrary saying x86 emulation on the M1 is fast enough even for demanding video games.
It truly is a shame we have to choose between quality hardware and a quality OS.
My maxed out 2013 Macbook Pro (the last model that runs linux without issue) has finally worn itself out and I’m in the market for a new laptop, but I can’t find a new laptop worth spending money on. I’m currently doing all my work on my desktop just hoping I don’t have to travel any time soon.
Look at Framework. Really fantastic out of the box Linux support (some WiFi issues depending on kernel version but legit it’s better than what you’ll find on identically specced Dell's right now and you have options about brining your own preferred WiFi card), excellent Windows support, a good community, repairable — I’ve had mine since September and am a huge fan.
I think you convinced me to not get a new MBP. These are the exact kind of issues I had with prior MBPs. Also Emacs was always inexplicably 5x slower than on Linux. It's a shame we can't use alternative OSs on the hardware.
I'd swear I'm hearing myself talking about Windows 20 years back... "I keep saving everything incase it just crashes", "It's slow as hell even for simple tasks", etc.
Has Apple 2021 become Microsoft 2001 ?
What you describe is not a Mac Mini with some OS or HW bug, it's a broken one. Don't know what your issue is, if it is hardware, or it was pwned, or has 2-3 incompatible Rosetta apps wrecking havoc. But I'd get it replaced.
None of these things happen to people, much less altogether.
>Everything runs slower on MacOS. Every. Thing.
The experience of everybody who bought an M1 is everything (or almost) running far snappier.
>Everything runs slower on MacOS. Every. Thing. grep, sed, awk.
So, is it "everything" or is it "grep, sed, awk"? And what is your case exactly? Benchmarked them on the mac and on some Linux system with the same file? Maybe did it on a network attached drive?
>Why don't sed, awk, grep, etc.. all the Linux things work quite right? I need to add something else on macOS to get them to work like they did on Linux.
You seem not to understand this POSIX system or versioning. Apple has older (or BSD) versions of those tools. Use Brew or similar and install the GNU versions, which would be like you know them from Linux. Not that the BSD versions are broken, they're just different.
Expecting the GNU and BSD apps to be identical is certainly understandable, but explaining that they are different is hardly the same as “you’re holding it wrong.” It’s information. If, instead of learning from it, you want to be mad at that information, let alone the messenger, I pity you.
> What you describe is not a Mac Mini with some OS or HW bug, it's a broken one. Don't know what your issue is, if it is hardware, or it was pwned, or has 2-3 incompatible Rosetta apps wrecking havoc. But I'd get it replaced.
> None of these things happen to people, much less altogether.
> The experience of everybody who bought an M1 is everything (or almost) running far snappier.
Especially since there are reports of the same thing as described happening qll over the comments here.
I did not give due credit to the rest of the comment, it was quite well written. I apologize for that coldtea.
Putting aside your dislike for BSD versions of userspace tools, I am confused as to why it hasn’t occurred to you that you may have defective hardware or a defective OS install? Have you undertaken even the most basic troubleshooting steps?
A lot of those weird random things would happen to me on Linux. But I agree with you, I wouldn’t expect a $2K+ vertically integrated, 10th iteration flagship laptop to have these kind of issues. It definitely does suck. I have the 16” base model M1 Pro and I’m also considering a return
> Why don't sed, awk, grep, etc.. all the Linux things work quite right? I need to add something else on macOS to get them to work like they did on Linux.
This is the GNU in GNU/Linux :)
I really like GNU coreutils, and other basic GNU utilities like GNU sed, find, and awk. I know some BSD folks see them as ‘bloated’ and ‘non-standard’, but to me they're plenty fast and they feel complete.
When I used macOS for a while for a job, it surprised me how much I'd come to rely on the features and behavior of certain GNU utilities that are apparently peculiar to them. But I miss 'em when they're gone.
I still don't understand why the majority of people haven't switched to Ubuntu or Kubuntu... They're quite user-friendly nowadays and they perform amazingly well. I often go for months without restarting my Kubuntu laptop. It performs exactly the same as it did years ago when I first bought it. What I like most about Linux though is that it's open source. It actually feels like my laptop.
I have Windows installed on it as a dual-boot and I cannot believe it when I log into Windows (Windows is definitely worse than Apple)... I just don't understand WTF happened over the past decade; Windows seems to have gone backwards. There is no way in the world I could ever use this thing and be productive. The interface is so clunky with title bars which use up all the valuable screen real estate. Difficult to switch between different windows. The cursor was almost impossible to use with the touchpad (it just feels unresponsive); astonishingly, I was actually able to get the touchpad working well with Kubuntu with the generic open source touchpad driver which came with it (it's extremely customizable).
Somehow, the narratives around open source software being inferior quality always seem to be disproved in the long run. While the quality of enterprise software tends to degrade overtime (as big corporation financial politics and bureaucracy creep in). Open source projects are free from toxic financial incentives so they tend to keep improving over time.
I love my Regolith/Ubuntu setup and don't want to move anywhere.
However, I can't find a decent machine which will last for 4 5 hours on battery, and it is important. My brand new Thinkpad T14 lasts less than 3 hours, and I am on minimalistic setup with only few apps running. Tried all kinds of power management tweaks.
OTHER thing is the issue with the microphone on external devices in Linux which is pain, which is why I always have to use my phone for calls.
"I write for a magazine and web site called MacWorld, yet I'm unable, unwilling, or both, to really look in to the problem. Let me complain and hope someone from Apple reaches out."
Maybe there is more than one issue here, but Intel users having the same issue on Monterey where they did not have it before seems telling. Maybe M1 machines had a different OS build or features that caused this issue in Big Sur and now it's universal in Monterey?
This may sound random, just my gut feeling for debugging kicking in. Since Control Center seems to be implicated in this, and Control Center opens ports for Airplay connectivity[1], has anyone tried turning off Airplay sharing? Or just turning off wifi/connectivity completely and seeing if the memory pressure lets up?
My M1 Air gets a ton of abuse and I don't see any abnormal perf issues. There was a super annoying bug where the laptop would crash when I closed the lid with a bluetooth mouse connected, but that's it. I'm still on Big Sur, maybe Monterey is buggy.
Apple has bunny hopped hardware architectures, successfully. The time is right to switch from the 70s FreeBSD software environment to Bell Labs's 9P concept-oriented operating system jazzed up with Apple's delightfully intuitive UX Pixar styled art and animation. Linux is a step sideways from the BSDs.
If such an issue were to occur on iOS without a crash, would anyone even notice? There's no real way to monitor RAM usage, and the failure mode is less clear (slowdown, issue with responsiveness, followed by likely reboot).
iOS would likely kill the app as soon as it got out of hand, and rely on the state restoration hook to bring back the state as the user left it. Normal macOS apps don't have that, though.
If the leak was in some part of the system, maybe a problem, but I'm sure they test pretty rigorously, and get enough crash reports to figure things out fast with that many devices in the wild, all in pretty constant use.
It may have to do with many of the new Mac (Apple-authored) apps being Catalyst apps. Not sure. I feel that Catalyst is a bit of a "stopgap" solution, so it may not be as thoroughly optimized as a more permanent solution. I'll bet something new comes out, once the entire line is switched to MX chips.
For myself, I just got the new M1Max 14-inch MBP, and it's quite nice.
I had to remove my Thunderbolt docks (I have a range of docks, including ones purported to be "Thunderbolt 4") from the video chain. I have found that the computer seems to lose track of what is where, if I have a dock in the loop.
SwiftUI is still better called beta software at this point, to the point where 1password abandoned it mid rewrite to go to electron. It's gonna be a few more years at least.
Considering that many corporations are just, now, beginning to take up Swift, after it has been out for seven years, I'd say this was a pretty sure bet.
I like SwiftUI. It shows a lot of promise, and I look forward to releasing products, written in it. I don't think it's especially mature, and maturity is important, when you need to release ambitious projects.
I'm not really in a hurry, though. I'll be getting around to working in SwiftUI, and I tend to do well at the things to which I apply myself.
The browser taking up so much memory are probably just object/memory leaking advertising scripts. Run an adblocker and watch the memory usage of your browser tabs go down significantly!
Nope. As pointed out, Firefox had a nasty bug that was interacting with an Apple bug that causes memory leaks. Firefox has fixed the bug but from their comments it seems that it was more of a workaround around the Apple bug rather than a fix.
I also do this, but I wonder is it a bad idea? Is Activity Monitor enabling any extra logging that itself is slowing things down further, or is it just subscribing to already enabled events?
I'm on M1 MBA 16GB and I get the out-of-memory errors too. It's still a world of difference from the 2016 MBP I upgraded from.
As touched on in another comment, the problem might be the 8 desktops and 20 Firefox windows I keep open.
I've always tried to avoid keeping Activity Monitor open. On my current 2015 MBA, it doesn't eat up a lot of memory, but it usually does list itself as a top consumer of CPU cycles, often 15-20%. Running `top` in the terminal only eats about 2%.
It would be ironic if Activity Monitor itself were causing a memory leak.
I'm a long-time Windows user and it's customary to keep Task Manager open (more for its tray icon) to keep an eye on what the system is doing. Is that not common in the Mac world?
We have apps like istat menus that behave similarly. Activity Monitor doesn't create a menubar icon that lets you monitor system state like the windows equivalent, but it's dock/taskbar icon can be set to behave the same.
I also think windows task manager is fairly efficient and TBH more useful with their full system state graph view. You have to click one resource at a time to see system state of your memory, cpu, etc and there is no GPU graph like windows.
Opening it from time to time, sure, keeping it open, no. Power users may install a menu bar applet to track the stuff but you generally don't need to bother.
Honestly unless something seems to be sweating the box there's not a great reason to monitor it all the time. By far the biggest CPU/memory culprits are web pages, but just looking at a bunch of chrome or ff processes doesn't really help determine which it is. Activity Monitor's own GUI is a pretty serious offender, running around trying to allocate icons and graphs for all those apps. If you just need to know what's going on, `top` on the command line will watch it for you. But usually (from a 2015 mac perspective) there's no need to keep an eye on it.
I did notice that CalendarAgent uses around 350MB which is ridiculous to just sync my Google Calendar. When you kill it, it restarts itself goes down to 15MB but gradually goes up. Doesn't seem to go above 400MBish, but still... Other than that, no complaints. Other apps seem to manage memory rather well.
Rather than working with Activity Monitor open 24/7, just install iStat Menus. Makes it really easy to monitor your system.
I have a MacBook Pro (13-inch, 2018, 2.3 GHz Quad-Core Intel Core i5, 16 GB 2133 MHz LPDDR3) and now on macOS Monterey everything just feels sluggish (especially the animations) and crashes sometimes.
It just feels the OS is designed for different, more recent hardware. I have been using Mac since 2010 and that wasn't the case in the past. I paid about $3,000 for this device. I am not happy and considering Linux/Windows.
Reading this with mild concern as I'm waiting for my first M1 Max, a couple questions spring to mind...
* I wonder if this is more prevalent when the hard drive is approaching full capacity? It would be interesting to see the breakdown between physical and swap when it happens.
* Has anyone observed these seemingly random memory leaks while not running anything under Rosetta 2?
I'm facing the similar issues with my m1 air(8 GB). Recently firefox managed to get to 44GB. I'm closing and re-opening my applications; and checking the memory pressure when things start to feel sluggish. Also unrelated but my speakers have started to sound like garbage.
Interesting hypothesis that it’s because MacOS doesn’t handle the memory structure well. I would have expected this to have manifested under either arch or neither.
These situations are interesting. My M1 13” will occasionally get a stuttering cursor that only goes away with a restart. I haven’t had this problem in over a decade on any computer.
I know I’m not unique in this because of the number of threads on the Internet on the issue.
Interesting to see the odd interactions between software and hardware. Though this sort of bug does put me in mind of how in Minecraft when Notch fixed light transmission through glass, it would start raining through it too.
Funny. The last time I had this was in 2014 or something when I had too little RAM and the system started swapping. Linux GUIs perform poorly under memory pressure. Way worse than windows or MacOS for some reason.
Been a constant experience for me from 2001 when I first started using.
My Intel-based iMac (32GB) has been struggling in Safari lately (more than usual) with stutters and general poor performance; there was a Safari update not too long ago, right?
Been using an M1 mac and Big Sur on the daily for a year and have never experienced this, and I have the 8GB model where I would notice it. My Control Center is using 29 megabytes.
I think this specific memory leak is only in Monterey. I definitely have to manually kill Control Center every few days otherwise it'll be using 20GB+ of RAM. Supposedly it's fixed in the 12.1 beta though, so hopefully Apple releases a fix for it soon.
The author specifically mentions that he is using macOS Big Sur: "That's extremely similar to the MacBook Pro and Monterey issues I’ve read this week — except I’m still running Big Sur."
It really highlights the value of the fleet-wide profiling that ChromeOS enjoys. There’s a dashboard somewhere of what’s using all the space (and time) for ChromeOS users in the wild. But HN just actually hates telemetry more than it hates memory leaks.
I'm running 12.0.1 and Control Center is using ~65 MB, with an uptime of 12 days. This is on an M1 MBA. What type of machine do you have? I wonder if there's some specific software that causes/exacerbates the issue.
What I find common on such threads is Linux users bashing how Apple users are suckers. I don't get the point of such comments. Most people on HN can already figure out how to build their own PC, install and configure the distro of their choice. But thats not even relevant as everyone is pointing out the flaws in software so that the company responsible will fix them. Bragging about Linux makes no sense, many of us using Macs here most likely already have a linux box as well.
I've been dealing with a weird memory leak on my Windows desktop for a out a year now. For some reason, dwm.exe (this seems to be the window manager) uses more and more dedicated GPU memory. This is really weird since it uses so much that apparently Windows does paging of GPU memory.
I recently got an m1 mini, just on principles the 16GB version, and did already get the RAM usage warning once, with my browser going into large memory usage. So there may be something to this story, will watch for it.
Huh I have an 8GB air and I run Kubernetes locally and develop a fairly complex system with Skaffold on it (~10 services). Thing is a beast, had no problems at all, but one does need to be aware one is not on the 128GB-of-RAM workstation :P
I mean how much demand will a locally hosted dev service generally have? Nothing about K8s inherently denotes high ram usage.
Browsers are generally the biggest culprits for memory usage. Random tabs will leak memory and before you know it some news article page is using 1GB, your GMail is using 1GB, etc.
I feel like this claim was made largely by nontechnical YouTube influencers and some Mac fanatics talking about how fast the swap on the SSDs is and equating that to not needing more RAM.
I don’t think it’s a wholly unfair assessment. You can get away with more when the memory management stack is smart and fast, but in the end if you need 9gb of ram and only have 8gb you’re going to know it. Most people just don’t (at one time).
Having this on my 13 inch Macbook Pro M1.
After upgrading to Monterey it also started to heat-up without any heavy app being open. In Big Sur I had to open say OpenTTD to make it warm. Now it's just browser and IDE doing that.
Is it specific to M1? On my Intel Macbook Pro I have to restart Safari daily to fix memory leaks that make the address bar incredibly unresponsive (even closing tabs doesn't work - I have to quit and reopen the app).
My hot reload JS dev server crashes on my M1 16GB between 1 to 7 times a week. This has never happened on my Windows 10 desktop machine (16GB too) with the same code base. My M1 is still ~15% faster though.
Managing shared memory is not a new thing for Apple, as it has been doing the same on iOS for a long time. Maybe a bug in Monterey? None of the M1 users I know had this issue.
> There are also lengthy Hacker News and Reddit threads with users recounting their experiences.
For many years, the press kept HN a secret and wouldn't even link or mention it as source. Presumably to keep HN from falling into mainstream. There were lots of times you could tell they got it from HN, or even screenshot without mentioning the site.
But for the past 2 years it seems more and more mainstream media are giving direct links.
I could only wish the first rule for mainstream media is to not talk about HN. But then I guess it is too late.
I think the way it caches media is bugged right out. I changed the cache size preference from 32gb to 8gb and it eased some of the pressure on swap, but it was still insanely high. I had to switch from the App Store version to the "Desktop" version you download from the website.
Perhaps a coincidence, but Telegram in a Firefox tab also uses a phenomenal amount of ever increasing swapspace. I've seen Firefox reach 67GB with a Telegram tab open, it is distinctly noticable compared to just about everything else I open in the browser.
Jumping in to point out that while this stuff happens on linux, at least one doesn't feel so hand-wrigingly powerless when it does. I know it's a choice between starting a career in maintenance and complete learned helplessness.. and the former option well seem a lot more appealing when it's no longer available.
I use Firefox on M1 Air with only 16GB of memory and have had zero memory issues. I have like 100 tabs open and no problems. Machine and OS works perfectly and is rock solid. No issues at all.
I haven’t had any issues. I sometimes run two VMs at a time on my 13” MBP with 16GB RAM and although the memory usage is almost maxed most of the time it’s been running fine.
M1 MacBook Pro (13 inch) and M1 MacBook Air were released near the end of last year. The newest machines (14 and 16 inch) and the M1 Pro and M1 Max MacBook Pro. The combination of "MacBook Pro" and "M1 Pro CPU" can be confusing. The original poster appears to have gotten an original M1 CPU in the 13 MacBook Pro about 10 months ago from their employer.
And that is why I won't be buying one until at least 2025, and unless they also remove all on-device spyware and let me power it off for real without it broadcasting its own location when shut down.
Apple Silicon has a lower power draw than any comparable AMD or Intel chip.
I'd like to use one of those Mac Minis as a homeserver/homelab and macOS is the only thing standing in the way. Similarly, I doubt there is a non Apple laptop that matches the M1 Macbook Air's battery life to performance ratio, specially in that form factor.
I know Asahi linux exists but reverse engineered drivers will never be as effective/performant as official drivers. Porting drivers to Linux won't even be expensive for Apple. They refuse to do it just cause theyre Apple.
I don't think there is anything else where you can run that much GPU power on battery as well as wired in a form factor anywhere close and with integrated system/gpu memory. The AMD APUs haven't gotten close yet, but as things start moving to LPDDR5 maybe they will.
I personally haven’t been able to find another laptop with the same battery life, design, trackpad, and screen quality of mac hardware.
I bought an M1 max laptop and I’m almost tempted to go back to my Linux machine because of several issues I have with macOS. There are things that are by design like mandatory animation delays for things like switching spaces and going in/out of fullscreen, which are borderline dealbreakers for me given how often I do those things. And there are also things I suspect are either bugs or straight up performance lapses, such as scrolling through text in a terminal at 4k being slow if you have ligatures enabled.
I could go on, but I think it’s a perfectly valid sentiment (regardless of whether or not they were venting) to wish for mac hardware with a Linux OS.
You might find it helpful to turn on System Preferences > Accessibility > Display > [x] Reduce motion. I go through phases of preferring the fades it applies instead of certain animations, although it still doesn't get it to "instant". For terminal, iTerm offers better choices for you to determine the balance between perf and display settings, incl a GPU renderer.
Please don't post generic-indignant comments, or take HN threads into flamewar, or call names in comments here. That all pushes discussion in the direction we're trying to avoid here.
Memory leaks are a feature of a software, not hardware. I read the article and while the Apple and Adobe software can be blamed, there is nothing in this post that it would pin this on the new hardware and the author does not seem to be technical enough to understand this.
Totally agree. I read the article as, when I am on my M1 I often experience memory pressure related issues. Not, because of my M1 I often experience memory pressure related issues.
It's very likely future software updates can help address these issues.
Honestly, the machines haven't had wide deployment in the field for all that long for Applications as huge as Photoshop. It takes time to discover, fix, and redeploy things like Photoshop. And it's not like Photoshop never leaked memory on an x86.
It's obvious you haven't read the article and are reacting to the headline. In fact, the author highlights that the issue is probably with the system software (macOS):
That’s extremely similar to the MacBook Pro and Monterey issues I’ve read this week—except I’m still running Big Sur ... It’s possible that macOS isn’t managing this unified memory structure properly, and will continue to allocate RAM beyond what is available without freeing up RAM that is no longer needed. This is commonly referred to as a “memory leak.” Performance gradually deteriorates until you need to either wait for the RAM to clear, force-quit the app, or restart the machine ... it’s entirely possible that Apple fixes it in a future version of macOS without ever addressing the issue publicly. That’s what happened with the excessive SSD usage earlier this year, though that Apple says that was a “data reporting error” and not an actual problem. The memory bug is an actual issue that needs to be fixed as soon as possible.
The big thing is that M1 chips have RAM in the same package, which allows for performance improvement if the software takes advantage of the new architecture. It turns out that the system they came up with offers good performance but doesn't tolerate memory leaks very well.
It is made worse by the fact M1 MacBooks are limited to 16 GB of RAM, non upgradable of course.
So yes, it is a software problem, and it will get better, but the M1 is particularly sensitive to it.
It is just a wild guess based on the article, but the unified architecture is mentioned. So maybe each subsystem (system RAM and GPU in particular) all have direct access and therefore more of a tendency to fight for resources and maybe also complex and therefore more error prone locking patterns over shared memory.
Why doesn’t it tolerate memory leaks very well? The fact that the system remains responsive and able to close the application indicates that it does tolerate it well. I’d say macOS in general tolerates memory leaks much better than Linux does, which if you don’t have a swap disk and you run out of memory the entire system just stops functioning.
can some mod do something about this user "dang"? he seems to crapflood every discussion I try to read with these stupid judgemental comments and never contributes to the discussions, doesn't hacker news employ moderators of any sort? I know I'm a new user and don't have enough "karma" or whatever to downvote but this guy is one of the most irritating and useless commenters I've ever seen, and that comes from someone who has used reddit and slashdot for quite a bit.
One wonders if there is some weird software his employer installed? Probably not. Apple just dropped a new chip into a laptop designed for another chip and its not a good match. Mac controls the entire production process including software. That laptop was designed for a different chip. Everything else with the M1 is fine because the hardware was designed for the chip
I've been having some sort of severe performance issue in one form or another since Mojave/Catalina or so with a 2017 MBP and a 2020 M1 Mac Mini.
The symptoms is always generally poor performance after the system has been running a while (4h to a week, varies), usually with WindowServer using CPU cycles non-stop and UI that felt choppy across all programs.
This seemed to happen frequently after "opening many files", like doing some recompiling with Xcode for a few hours, or indexing a large volume with Spotlight. Rebooting helps temporarily.
Today I realized that data read/written since boot was about 1TB in a few hours on a brand new OS install, and I traced this back to the com.apple.Safari.History process. Somehow having bookmarks and previously using Safari 15.x caused a huge amount of I/O that wouldn't stop - the solution was to remove all bookmarks and reading list items. Performance was immediately back to normal, no reboot needed.
So just logging in with your iCloud id, you could be "importing" whatever performance problem you're having on a new install.
I recommend you reboot and take a look at your disk I/O stats - maybe this will help someone!
Thanks for this tip. I’ve had a similar problem occasionally and this gives me an idea for some things to check.
Have you looked at Keychain items as well? It seems to go very slow for me, and I wonder if that’s part of some similar-sounding slowdown I have.
Quitting Safari seems to resolve the slowness, but not always.
Yes, noticed this some time ago with WindowServer taking all my CPU usage but only twice so far for this year I think so it was okay. When it occurs every day then I will start to worry then
Might be https://apple.stackexchange.com/questions/407177/windowserve...?
IIRC, when the M1 first came out, there were a bunch of people saying their SSDs were being worn out super quickly, citing SMART statistics. Perhaps this safari process was the culprit?
> Perhaps this safari process was the culprit?
No, for many people it was Rosetta apps.
BTW, Apple totally lied when they said they "fixed it" in an update and it was only a "reporting issue". It's not fixed, and it was absolutely f'ing not a reporting issue. People's SSDs have already failed because of this, and obviously they're soldered.
> People's SSDs have already failed because of this, and obviously they're soldered.
Any links to more info about this? I’ve had my eye on a 16” Pro with an M1 Max for a couple weeks now but want to make sure it won’t have issues like this.
I haven't seen any conclusive evidence one way or the other.
I haven't actually seen any reports of SSDs actually dying. Can you provide some links? SSDs dying in less than a year of typical use is a huge deal.
https://tidbits.com/2021/05/27/an-m1-mac-cant-boot-from-an-e...
I don't see any reports but there is evidence that should the soldered on SSD fail the entire device is borked short of board level repair.
wow Safari is the IE5 of browsers now, having to do so many work arounds like we used to have to do with IE5 and now its killing hardware that you can't even replace.
IE never killed hardware. It kept legacy Enterprise technologie alive ( eg. ActiveX )
Chrome is IE of browsers now. 20+ years ago nobody cared about non-IE, today nobody cares about non-Chrome.
Chrome is the IE of browsers with respect to being the dominant player approaching monopoly status.
Safari is the IE of browsers with respect to being buggy and behind on features with a slower release schedule.
IE was both and more extreme on both issues.
It’s only “behind” relative to Chrome, the dominant browser which more or less sets the standards these days. Calling Safari “behind” is judging browsers by how similar to Chrome they are.
Also Firefox and even Edge when the IE version was being developed. They've also refused to implement features which would bring web apps in line with the the capabilities of native iOS apps (things like push notifications, offline support with data that doesn't get cleared every 7 days).
Unsure whether it's Safari related, but it was solved on a regular macOS update. I didn't recall any specific Safari update, but I might miss it.
Yes I've seen something similar. WindowServer taking up all the memory.
In one case I also had it crash while I was out, with every open program opened several times on the dock. I didn't do this and in fact macOS doesn't let you do this. Everything was hanging completely so I had to turn it off and on again :)
But this kind of thing does not instill a lot of confidence
> in fact macOS doesn't let you do this
See open(1), in particular the -n flag.
Do you run your laptop on non—native resolution? Or an external monitor with a non-native resolution?
I'm runnign on 2018 MBP, 16GB RAM + 4k external monitor. I experienced the same type of issues a couple of months back — high WindowServer CPU, mega choppy UI after a few days of use. Initially thought it was Safari, but it kept happening with other browsers as well. Researched it a bit and found a thread where someone suggested running both the laptop and the monitor on native resolution. Haven't had any problems since doing that. WindowServer sits at about 10% CPU and ~2GB RAM, current uptime 10 days.
I think I had the same problem as you (similar hw and setup). Do you have a discrete GPU? If so check the link below out. For me, this behavior was due to OSX flip-flopping between discrete and internal GPU. Once I set it to discrete all the time (I'm always plugged in) the problem went away.
https://support.apple.com/en-us/HT202043
Seems to me a different issue than what people are complaining about atm... but what do i know, might be one and the same.
Edit: very weird... Looks like this setting was reverted for me. I just updated from 11.? to 12.0.1, so I wonder if the installer undid some of my changes. No performance issues, but I've literally just my computer on for the first time after the upgrade.
Nope, I just have an integrated Intel GPU.
I'm also pretty much always plugged in. It's likely that the issue was manifesting before switching to a 4k monitor, but after switching the issue became unbearable. E.g I'd wait 2-3s to switch between windows or desktops. Running both screens on native resolutions and I'm switching between windows near-instantly.
¯\_(ツ)_/¯
:( Sorry to hear that... I know it must be so incredibly frustrating!
Sounds like it isn't an issue of flip/flopping between discrete and integrated GPU then... must just be the integrated GPU and its ability to deal with non-native resolutions.
SafaribookmarkSync, Cloudd, along with Content Cache has been problematic for years. Especially if you have huge number of Tabs and Bookmarks that is being synced across devices.
Along with iCloud syncing for one reason or another have relatively higher probability of being messed up during update, and you end up with a scenario where something is to probably synced and it keeps trying it over and over again.
I know you’re humble bragging with Lunar but I’ll bite.
Thanks for such a quality app.
Im only using it to fully turn off my built-in monitor while working in clamshell mode and then manually setting the external’s brightness.
So no sync mode (and not even sunset sunrise)
But solved my problems nicely and no more reaching out to my external monitor’s buttons in Narnia
If I’ll keep using it I’ll make sure to caffeinate you ;)
Count me in as another WindowServer casualty. Can't believe it's 2021 and MacOS can't handle a 4K external monitor without grinding to a halt...
I just ordered the M1 Max with 64GB RAM because I'm constantly getting the "Your system has run out of application memory." popup while working on Lunar[1] in Xcode.
Every time this happens, Xcode uses about 4GB of RAM (probably because of the monolithic UI storyboard of Lunar) but it should still leave enough memory for my other non-memory hungry apps.
But then I open Activity Monitor and I see WindowServer using ~80GB of memory [2]
The only remedy is either `killall WindowServer` or a full reboot.
I've been using an M1 MacBook Pro and Monterey since the first developer beta but this only became an issue in the last 2 months or so.
[1] https://github.com/alin23/Lunar
[2] https://cln.sh/CfEL6u
I’m probably dumb and wrong, but that 80GB is likely virtual memory and not a great indication of actual RAM used.
It is virtual memory, the real memory size is usually sub 1GB. But I just can’t find any other culprit, and killing WindowServer is the only thing that helps.
I have no idea how to troubleshoot this and don’t really have time for it when working on hotfixes, I just want to get back in a state where I’m not facing an imminent system lock up where I can’t even save my recent work.
Do you tint your mouse cursor (new Monterey accessibility setting)? I noticed this in the last handful of developer betas, and it definitely shipped out in 12.0.1: https://bugzilla.mozilla.org/show_bug.cgi?id=1735345
Although that particular issue will show in application footprint, not WindowServer. I see WindowServer buildup on one of my M1s still, even in 12.1 beta :(
Nope, never used that setting. I just opened Accessibility and hit reset on the pointer colors again just to be sure.
This is a windows technique, but presumably something similar might be available on mac too:
https://randomascii.wordpress.com/2018/02/11/zombie-processe...
Activity Monitor shows a "memory footprint", rather than just a simple sum of virtual memory usage. Among other things, this number splits memory usage for resources shared across multiple processes (e.g. system libraries) and attributes a fraction to each of them.
I’ve noticed the high memory usage of WindowServer on my Mac at work. To the point where I have to kill it, which is a pain.
This is an intel Mac though. I haven’t seen it on my M1 at home.
My god. I made it down to here in the thread wondering what all these people were doing with Windows Server.
Another anecdata: My old 2015 8GB MBP also has the same problem; unbearably slow now. The M1 MBA (also 8gb) doesn't have this problem.
I too have the problem with WindowServer although for me it seems to happen with just regular usage. I have not rebooted in 48 days, and currently it is using 10GB RAM (I have a MacBook Air with 16GB). I have seen some people claim it has to do with using a display scaling setting other than the default (I use 1280x800).
How many windows do you have open? 10 GB of RAM is a very large amount for WindowServer to use.
Yes, I agree, but it is a memory leak. It doesn’t START at 10GB. Also closing windows does not help much at all.
To answer your question, I have a total of 73 windows open at the moment. 11 of them are from applications with a single window open, 8 are from iTerm, and 54 are from Sublime Text. I am aware that is quite a lot for Sublime, but that is just how I use it.
Regardless, I just quit Sublime Text, and the memory usage only dropped to 8.9GB still absurdly high for having 19 windows open.
Yes, it is. Can you try running this and seeing what it reports: https://news.ycombinator.com/item?id=29147886?
Thanks. Output is here if you have any ideas.
https://gist.github.com/ccampbell/9a0118b0e3e607ec56a3a2cc59...
Hmm, that is interesting. You have lots of unaccounted for memory, which is not normal at all. Would you happen to have the full footprint output? Trying to see how the 200K individual VM_ALLOCATE regions are distributed in side.
I have been seeing all of the above problems on my 16GB M1 MacBook Air for probably at least 6 months now.
WindowServer has consistently been the main culprit but occasionally other apps like to chip in where they can too. Often that means there isn’t one single process which I can kill to reclaim enough memory to continue working without being promoted again or having MacOS forcibly close all my applications on me. As a result, reboots appear to be the only real solution for me.
Interestingly though, in the past few weeks I have been working with similar numbers of open Sublime Text windows as yourself and I can confirm that since then, ST4 has been a regular and serious offender when it comes to memory usage in Activity Monitor.
Great to see you pop up on HN. I discovered Lunar a couple of weeks ago after wondering whether something like it existed and it's been a great tool so far. Solves one thing and solves it well. Keep up the great work!
Thank you for the kind words! I’m always happy to see that the hard work that goes into this app is appreciated and useful.
Where do you see the 64GB option?
The choice becomes available if you select M1 Max as the CPU: https://cln.sh/A3TcaP
Ah, interesting, the page I wound up on did not have selectable elements.
That's what I find amazing with Apple users.
They sell you a very expensive machine that doesn't work.
Your reaction?
Buy a more expensive one from the same provider!
What other machine can you buy to run Xcode on?
Also, I get along just fine with the last few Macs I have had...
How about the Mac Mini?
Or Hackintosh. Hey, GP asked 'what machine'. Didn't require it to be an Apple.
I’ve given this thought for an older MBP of mine that can’t upgrade to latest OS so that it can be used as a CI.
My issue with that is 1. Can I trust the hackintosh to not be compromised and 2. Uploading an IPA via hackintosh feels like bananas.
If you use say Proxmox and create your own macOS image and compile your own Opencore then you should be OK. You can do same on VirtualBox. (On a PC, cause I doubt older MBPs have the necessary stuff like VT-d.)
Of course, once Apple kills off x86-64 macOS it is going to become more complicated.
Hackintosh's are dead as soon as Apple stops supporting Intel chips.
That won't be for a very long time. There's a lot of Mac Pros out there. They're still supporting the 8-year old 2013 Mac Pro, the trashcan one. They sold a lot of the revised Mac Pro to animation studios and such. I wouldn't count on support for those being dropped for at least ten years.
That would be the same for OPs use case. He’s using a M1 MacBook Pro so nothing would change. The M1 devices don’t support more than 16 GB RAM, too.
The recently released M1 MacBook pros go up to 64GB.
That seems like another symptom of the same problem. Apple locked me into their ecosystem...but that is their right, after all it is their platform. Apple sold me faulty hardware/software but I have to buy more Apple products (new hardware, support) since I am locked in and cannot do my work on non-Apple hardware. Their behavior may not be illegal (IANAL) but I find it quite user hostile. The main problem I see with all this is that since Apple is getting away with it, more and more companies are trying the same tactics and normalizing user lock-in and charging extra to fix faulty hardware and software.
As someone who hasn't been in the Mac ecosystem since his and his mom's iMac both broke in the same way in high school, how does the Mac ecosystem lock you in?
Comment just above mentions the need to use macs because of xcode.
As far as I know you need to use Xcode to distribute the software, or to write macOS-native software. I have written plenty of Java Swing apps without touching Xcode at all.
To be fair to the OP, the "more expensive" machine is also probably the fastest laptop on the planet at compiling code, so there's also that.
You can use alternative toolchains like like cctools-port[1] and xcbuild[2]. See, for example, this article [3] on how to build Swift UI apps for iOS using Linux.
[1] https://github.com/tpoechtrager/cctools-port
[2] https://github.com/apple-cross-toolchain/xcbuild
[3] https://thi.im/posts/cross-compiling-for-ios-part-1-build-sw...
Wow, they make it so easy!/s
Amen. I can't think of another brand that fosters so much "I'll spend more money here" compared to "I'll use a different brand."
That's how much we hate Windows and desktop Linux.
Good thing you can afford the kind of expense that the luxury of that hate is costing you then, I guess.
For a professional developer the difference in cost is not that much bigger.
Well, generally speaking Windows and Mac developers are in such demand that hardware is quite a small expense compared to how much they make.
Is the experience very different for Linux developers?
It's very weird to be talking about "windows, mac or linux" developers.
Desktop application development today is maybe about 1-5% of all development, while overwhelming majority of backend development - the largest category - targets Linux.
Percentage of the way to being as good as all desktop operating systems ought, at a minimum, to be:
Windows: 70%
Linux: 20-80% depending on your hardware, what you're doing with it, and how lucky you got with the current kernel & libraries your distro is running.
MacOS: 90%
So when Apple fucks up really bad and MacOS temporarily drops to 70%, no, I still don't switch. Every desktop operating system is shitty, but there are degrees of shitty.
And yes, I use all three pretty regularly.
Crashing while doing your work is not "drops to 70%". That is the problem we were discussing and that is not a minor inconvenience.
Oh look the main speaker for the Mac folk!
Probably because other brands don’t do anything novel to warrant such loyalty. Apples hardware and software stack is undeniably unique which in turn makes it a brand that people will put up with issues because well, it has things other systems don’t.
These issues notwithstanding, the total user experience of owning a Mac laptop is far superior to any Linux or Windows machine I've tried, at least in the past 3 years or so.
Two concrete examples: trackpad feels like shit on both Windows and Linux, installing or updating drivers and some apps on Windows constantly switches back to legacy modes to display the installers (with different fonts, no subpixel smoothing etc.). If it's not a deal breaker, it is annoying as hell and something that almost never happens on my Mac.
libinput with Wayland on a MacBook Pro puts its trackpad handling on par with macOS.
I'm dailying Wayland on Ubuntu, but it feels like there's always a "it'll be really good soon" promise to Wayland.
yea it takes time but eventually open software will be better, just look at how well Blender is doing and even Darktable is way way better then Adobe Lightroom can't understand why anyone uses that anymore while a free alternative that is better is out there.
^ This. It's not that I don't want to switch to a different company, it's just everything else is a worse, overall experience. I'm on Windows every week or so, and while it feels much faster than Mac, on the same hardware, there are just too many little things that annoy me. (have tried a few Linux flavours too, not really a contender for my use case)
OP was describing very high-spec hardware which locks up and becomes unresponsive under their normal workload, potentially losing work or wasting their time. That seems like a much more significant issue than inconsistent font display or the feel of the input device.
Your experience must differ, but if I spent $2000+ on a device only to find that was locking up (due to poor memory management in a standard component!) under my standard workload, I would be livid. I would probably return it, and then tell the next 10 people I meet about my poor experience.
That said, I have run across a Windows 7 bug where the font manager would repeatedly fail and leave its cache/log on disk, only to restart and repeat until the hard disk was just about full. Complaints about the poor state of software engineering as a discipline go back to the beginning.
As a counterpoint, I was a Linux user for ~20 years (eventually with a macOS dev machine on my desk as well).
I got so sick of troubleshooting issues myself, I switched to Windows 7 and have never had to troubleshoot my OS since then. I don't google Windows issues, I don't have crashes, nothing.
It's nice to have a machine that I can just work on and not constantly debug.
(I 100% expect privacy advocates to admonish me for this, but I use hardware I assembled myself and block MS telemetry.)
In a world without MacOS, I'd probably be running Linux on a really beefy Windows machine under virtualization, for work. Desktop Linux is so much more stable and pleasant to use in a VM with Windows or MacOS handling the actual hardware. Still a bit unstable, but not nearly as bad.
> These issues notwithstanding, the total user experience of owning a Mac laptop is far superior to any Linux or Windows machine I've tried, at least in the past 3 years or so.
That.
I don't love Apple. They're a corporation, so at best our interests temporarily align. Loving them would be absurd. I really really want them to have strong competitors. I'd love to be back on an open source OS full-time, like I was for many years.
Unfortunately, they're so far ahead of the competition, that they screw up, sometimes even in a few ways at the same time, and people come out saying "LOL and Apple fanboys won't switch to Linux even now", and that's true... but it's because I'd be trading a few problems for a few score problems.
I ran Linux on laptops and desktops for about a decade, as my main computers. Ubuntu near the end, Gentoo for about five years, Mandrake really early on, a little time with Fedora somewhere in there, a sprinkling of Debian. I still try it out every year or two! I wish it weren't, but it's still much, much worse, and in the best case slows me down and gets in my way more than a "bad" Apple machine with a "bad" version of an OS X / MacOS, barring actual faulty hardware.
Because Windows is so much worse. I'm now required to use Windows for a project, and it is a pain to use that system every extra minute.
You might get a better experience running Windows in a VM. Reports are that it is faster that way.
It's because they do so many anti-consumer things to lock their victims in.
Changing brands becomes more of an ordeal than staying.
Tesla. So many people seem to have problems with the cars/lack of quality only to say “still love the car” or “it’s my xxx number Tesla and they’re still doing [problem]”
It’s really something, isn’t it? I’ve never regretted my decision to switch from macOS to Linux over a decade ago. Having to restart a 64GB machine constantly because it runs out of memory? Meanwhile my 4GB Linux computer shows a 38-day uptime (only because I accidentally unplugged it last month) and my two Linux servers have uptimes of over a year.
Just to be clear: This perfectly fine experience you describe is also exactly the experience of 99% of Apple users, including me. You only hear the ones that complain.
The ones that don’t complain have low standards. When I switched, although OSX was pretty solid, it was trading frequently spinning beachballs for the first consistently responsive computer I’d ever used. I had come to regard macOS sluggishness as normal, as do most Mac users, because they haven’t experienced anything better.
Is this sort of sweeping generalization really necessary? I can tell you that it's not true in any case. I'm one of those Mac users that doesn't complain about performance, because I rarely see the spinning beach ball.
Lol. Sorry your own experience was bad. Doesn't mean everyone had the same experience as you though.
I ran windows, then linux for ~4 years as a daily driver, switched to mac about a year ago.
I do back end/front end development full time for work + work on many performance sensitive personal projects, and couldn't be happier with the OS.
It's almost like different users have different experiences based on their use cases...
I haven't seen a spinning beachball on MacOS in years. I'm not even sure it still exists anymore. But also only rarely have a sluggish system, with or without a spinning beachball.
what are spinning beach balls? i don't understand? like BSOD ?
A cursor indicating the frontmost application is unresponsive.
Definitely still exists. I get it in emacs sometimes (god only knows why).
Good to know! I definitely have not seen it in years! (two 2015 Macbooks running Catalina with 16GB).
That doesn't mean linux is any better. I built a brand new ryzen x5800 machine late last year, running linux mint. When I first built it, the computer has had duplicate unremovable mouse cursors appearing on the screen. And my trackpad sometimes randomly became unresponsive (or one time, the up and down directions flipped!). And neither suspending nor hybernating worked - the machine wouldn't come back up.
All these problems have since gone away thanks to software updates and (importantly) a BIOS update. It turned out my USB controller wasn't being configured correctly by the motherboard. If I didn't happen to check for BIOS updates, I'd still have lots of problems.
And that was still a better experience than the HP laptop I bought a couple of years ago. That laptop had so much HP crap preinstalled that it sat on 30% CPU usage at all times out of the box. The biggest offender was some HP process doing microphone noise reduction. It ran all the time regardless of whether or not the microphone was being used - and utterly demolished the machine's battery life.
Apple is far from perfect, but they have a better reputation than most vendors for a reason. And their technical support is are excellent. I've had thousands of dollars of free hardware repairs & replacements over the years when things have broken in my apple machines. I wish they didn't break so much; and I wish they prioritised fixing software bugs over adding features internally. (APFS aside, I would trade the last 5 years of macOS features for better stability and performance in a heartbeat.) But there's a reason they're so well liked on HN. They're still pretty excellent in comparison to the competition.
How well does macOS run on your home-built Ryzen machine?
Huh? What point are you trying to make? I don't care how well macOS runs on my machine. I care how well any given computer gets out of the way and lets me get my work done. Today that means intellij, the rust compiler, firefox and nodejs. I want to spend my day working. Not futzing around with HP bloatware. Not rebooting due to apple memory leaks. And not guessing why my keyboard needs to be reconnected.
In your story, you built a machine, and Linux didn’t work well on it at first. Later, it worked fine. Meanwhile, macOS never did and probably never will run on it. And you introduce this story as an illustration of how “Linux isn’t any better.” So, to borrow your words: Huh? What point are you trying to make?
In my story linux didn't work well at first. Then with some time and work, it worked fine. An HP windows laptop didn't work well at first - but after I spent some time removing the HP bloatware, it worked "fine". This article says that mac laptops aren't working well because of memory leaks. But in a few months Apple will probably fix their software bugs and they'll work fine too.
You said:
> The [mac users] that don’t complain have low standards
I agree with you. I'd go even further and say most users have low standards, because the out-of-the-box experience with most modern computers is pretty bad. Apple users should absolutely complain more when the out-of-the-box experience with their computer isn't perfect. We all should.
Apple customers keep buying apple computers despite their issues not because we have low standards. We do it because the alternatives are even worse.
“most users have low standards”
I think you’re probably right about this. It’s like we’ve been trained to have low expectations.
As an avid linux user I can tell you. Linux users aren't "trained" to have low expectations. They are trained to be masochists that enjoy pain. The raw amount of configuration, fixes and setup work that has to go into making some linux distros work is astronomical.
If nonlinux users are trained to have low expectations, Linux users are trained to rape themselves constantly.
I see your point of view a lot but I don't think configuring a system is that hard and I think it's preferable, compared to have tons of complexity preconfigured without your knowledge.
Here's my experience with Arch Linux: 1) You install what you need to install 2) You read the doc and configure to your liking, learning what the system can and cannot do and how it works with everything else 3) Things generally work as documented and you go your merry way 4) When something break or doesn't work, you have the knowledge or a general idea of what to fix
Compared to proprietary systems where there is no documentation and little configuration, when things break or misbehave you're essentially screwed. Will you wait for an official fix or look online for reverse engineers? You bought a black box and you can't just open it and fix it.
I'd rather spend a day learning about my system during the installation and having it work for years, instead of buying yet another black box I can't open.
>I'd rather spend a day learning about my system during the installation and having it work for years, instead of buying yet another black box I can't open.
Learning linux doesn't take a day. Second off those black boxes statistically are more reliable than linux EVEN when you account for all the things you talked about.
The reason for this is simple. Microsoft has a business advantage and unfairly strong arms hardware manufacturers to make their stuff work with windows. This unfair advantage makes the windows user experience better and more reliable than GNU linux whos' developers constantly have to play catch up with hardware.
It's is sad isn't it?
In the last 20 years PC hardware has become ridiculously faster - and yet the user experience seems to have changed little.
I saw an article ages ago whereby Microsoft apparently had a dependencies approval gateway. If you made any change based on Windows you had to remove/reduce your dependencies or make a good case for not doing so.
I was hoping that would catch on.
The point he's trying to make is linux on a Ryzen isn't as good as MacOS on an M1. The user experience is inferior. That is his point and his point is fact. Hope you understand now.
user experience is inferior? how so? what nonsense, i have heard countless stories on dumb things Apple forces on its users while on Linux you can just use Vulkan... Imagine if your app can't contact Apple servers they don't launch or take minutes to open...and when they do work they are recording information... why on earth do people still use closed source software. Now we have to deal with Safari crap... like we used to deal with IE5, having to have workarounds for browser standards Apple refuses to follow or implement.
Don't waste your breath, let them be 'happy' in their little cult. I think Apple users are often mad because subconsciously they know they are taken a fool.
Feels super ironic you choose to describe Apple users as a "cult" in a reply to a guy that sneaks his "Safari is the new IE5" in every response... ;)
It's not a cult. I use all three operating systems and I'm highly proficient using linux. My distro of choice is Manjaro and NixOS and I have a lot of experience with debian and ubuntu as well.
Each operating system has benefits and downsides. For user experience linux is definitively the worst. Definitively. The differences are so large that it's almost comical how someone on this planet can't see the difference. If you can't see this then you are the one that is part of a cult.
> user experience is inferior? how so?
I use linux every day for work and:
- Hidpi support is spotty (its getting better but still a bit spotty)
- Smooth scrolling in applications is inconsistent at best
- The keyboard shortcuts for moving the text cursor around are inconsistent between programs
- Hardware support is much more of a crap shoot. Linux has a harder job than macos in trying to support every combination of janky hardware out there. But as a user, I don't really care. I just know that if I buy a mac, the OS will work perfectly with the hardware on offer. That isn't true on linux.
- Lots of useful software isn't available on linux. Eg, I love Monodraw, but thats macos only.
- App distribution on Linux is a mess. Apt? RPM? Snap? Flatpack? Etc etc. I have 2 copies of discord installed for some reason, and I have no idea what the difference is between them.
Etcetera..
Linux gives you the choice and freedom to spend an unlimited number of hours customizing everything. With linux I'm in complete control and I love that. On macos, things usually just work out of the box. I love that too.
No OS is perfect. There's tradeoffs with everything. If you don't understand other people's preferences, that doesn't make you right. It makes you ignorant.
Linux can't even run on the M1.
He's obviously comparing a linux PC experience with an apple OSX experience. It's a completely valid comparison but you're just being mean an inappropriate.
Linux does run on the M1. Not great, yet, but it's getting there.
My bad. You're right. So let's compare Linux on the M1 with OSX on the M1. Which one has the better user experience?
Apparently it runs just fine a little fiddling. https://github.com/mikigal/ryzen-hackintosh
Tell that to my tech lead who spent more than 8k buying and repairing his macbook due to constant motherboard issues. I love macs but don't dare to try them due to soooo many people in my circles seem to have very frequent HW issues with them. It seems to be either in the shop being repaired or having issues with connecting bluetooth stuff. I'm used to Thinkpad reliability so not looking forward to trade that up for some pretty pixels. My lenovo laptops had 0 issues in the past ten years. Zero. But now I'm web development and I need a mac for Safari to test locally.
I manage a fleet of macs. I have not seen any issues like that. I'd ask what your friends are doing to their computers to be seeing such a disproportionate amount of hardware issues.
> The ones that don’t complain have low standards.
Wow, well, thanks a lot. I've been developing on both the Mac and Linux for years and I don't have this issue on any of the multiple MBPs I use, including my personal 2013 model. I'm not saying I have no problems on the Mac, but they're not worse than the issues I had with Linux, especially on laptops.
I switched to a 13" M1 after having a Dell XPS 15 as my daily driver. I had high standards and Apple blew them out of the water. I see some beach ball on occasion, but I slam the machine hard. (8gb ram). It has been more reliable than the XPS 15.
Better than a Dell XPS 15 is no standard at all. The Dell XPS 15 is an utterly terrible laptop that breaks if you close the lid and put it in your bag when you get up from your seat. It's a portable desktop falsely advertised as a laptop due to its many flaws. If the best that can be said for apple is better than Dell XPS 15, that is a really, really damning review. I hope it's vastly better than that for you.
> Better than a Dell XPS 15 is no standard at all.
so what is the laptop that would meet this standard for which you're making a judgement?
So far, apple hardware has been superior to any other manufacturer, even with their warts, and their anti-consumer practises (of not having replaceable batteries, etc).
Not OP, but I have to chime in here- Dell's XPS line somehow got a reputation that it absolutely did not deserve as a good laptop to run linux on (or a good laptop at all).
I've been perfectly happy with a surface book 2, and various Lenovo and HP laptops. They each have better and worse product lines. Currently rocking an HP elitebook for personal use and a MB Pro for work.
The hardware quality is more or less the same on both. With the pro, you get a moderately better touchpad, a bit thinner and metal chassis. On the other hand, crap keyboard and the abomination that is the touch bar, and no legacy ports.
Personally, I don't think I would buy Apple for myself ever again, but that's just me.
The fact that we even have to debate about which laptop is good for Linux or not makes me not want to bother. I switched to Mac 12 years ago, and I never had to bother thinking about choosing hardware since. It just works.
It got the reputation because it is well deserved. The clickpad is far superior than macs and the software support is great
No one will ever convince me that a laptop with a web cam whose position is limited to looking straight up your nose is quality.
I primarily navigate by keyboard, so the difference between an acceptable and amazing trackpad is pretty minimal for my preferences. On the other hand, port wonkiness, temperamental wifi, okay keys, mediocre battery, driver issues and an utterly useless webcam convinced me to ditch the XPS the first chance I got. It's amazing I used it nearly 18 months.
> Dell's XPS line somehow got a reputation that it absolutely did not deserve as a good laptop to run linux on
This is where it started... https://www.engadget.com/2009-05-24-dell-now-offering-studio...
>So far, apple hardware has been superior to any other manufacturer, even with their warts, and their anti-consumer practises (of not having replaceable batteries, etc).
Milage may vary with other manufacturers, it doesn't look like you would change to another manufacturer. Which is ok.
Check business notebooks, from HP or Lenovo. (I forgot one particular Dell series.)
Ok... so what's the laptop that can compete with the new 14 and 16 Macbook Pros then and doesn't have some stupid flaw? The issue with waking up in bags seems to be caused by Modern Standby which is on most new Windows laptops?
I was going to say just disable modern standby because that was totally a thing until spring 2020 when the registry key to disable it was removed.
Microsoft is basically a toy OS run by incompetents without reasonable QA unfit to be anything but a launcher for stream.
HP Z-book, some Lenovo Thinkpads and some consumer Lenovo and HP notebooks.
I would go with a Lenovo or HP machine before a Dell. Even if they all have the sleep issue (I don't know if that's the case), both Lenovo and HP have better build quality.
I've had multiple Asus laptops (UX305, UX330 and most recently S435UA) and all of them ran Fedora flawlessly out of the box. Granted, none of them had Nvidia GPUs which is the usual pain point of desktop Linux.
I have had the opposite experience. Switched from a MB Pro that kept getting progressively worse over the course of a year (keys giving, overheating, lock ups) requiring rather annoying applecare interactions to an XPS 15. Granted it is not quite as pretty look at, but blows the mbpro out of the water performance and stability wise.
Thanks for telling me that I have low standards. Amazing considering MacBooks on my desk.
> "The ones that don't complain have low standards.... most Mac users... haven’t experienced anything better."
That's a pretty broad brush you're painting with right there.
It's in reply to "99% of all Mac users are happy" - which brush is more broad?
"100% of the 1%" is less nuanced than "99%".
Who am I to deny somebody the opportunity to feel superior? It’s a form of community service.
> The ones that don’t complain have low standards.
> they haven’t experienced anything better.
I don't want to gratify this pointless flame any more than to say that I have used all of Windows (XP, Vista, 8, 8.1, 10), Debian, Fedora, Ubuntu, and macOS as my daily driver at one point or another, and I currently use macOS because I have yet to experience anything better. There is a simple explanation for your problems here: something is wrong with your system (or you are overtaxing it) if it beachballs constantly. This is not typical macOS behavior, especially not since they switched to SSDs. I actually haven’t seen a beachball in years.
I don't disagree with you, but you didn't exactly refute the point of 'low standards' just by using multiple OSes ;)
My standards are high enough that I choose to use macOS and not Linux or Windows. I make no other claims about them.
More like "specific enough" instead of "high enough". Use Apple all you want but don't lie to yourself by saying Windows isn't a better, more powerful, more stable operating system.
It just isn't as polished, and even that's debatable.
I work daily in Linux, OSX, and Windows. If there was significant differences in responsiveness for day to day usage, I would notice.
OSX is plenty responsive, and I basically never see beachballs.
It's not a matter of low standards.
You are confusing your own personal experience for the standard experience of macOS users. Personally, beachballs are so rare for me in the past few years that I'm always amazed when I see one.
I'm sure I've gotten 100-days uptimes on my MacBook. Close the lid at night, it goes to sleep, open it up in the morning, and continue working.
I can’t get more than upper 30’s before it just starts doing too many odd things to work around
Yep, same for me. But the restart fixes it every time. That's why I set a timer that reminds me to restart after 20+ days.
Only if you skip updates that require a reboot?
Same here. I get ridiculous uptimes on mine. (And, yes, I defer SW updates…)
This has been my experience as well. I've had an M1 MacBook Pro since they came out and I've never had to reboot my machine outside of software updates that required me to. It has always just run.
I use my computer to write Elixir code, write music with Logic and Studio One, and edit photos/videos. Maybe none of those tasks trigger the problems other people are having.
Refreshing just HN probably doesn't take too many resources.
Don't OS upgrades come out more often than that?
I don’t know what you’re doing on a 4GB machine but I promise most workloads that create significant memory pressure would start creating problems for you, just like on macOS.
I use a dual xeon, 64GB Linux desktop which performs reasonably well.
A colleague runs a Linux machine with 16GB of ram. Compiling the same project can cause OOM issues for him and the kernel starts whacking processes.
Even if he closes all non-essential processes, his builds run 10x as long as mine due to memory constraints.
My main point here - let’s not pretend your 4GB machine is being used to do any intense local work.
The inverse of this is seen on my Intel MacBook Air and my colleague’s M1 MacBook Air. The compilation time for our code is 35 minutes or more on my machine and only 3 minutes on the M1.
> My main point here - let’s not pretend your 4GB machine is being used to do any intense local work.
That depends entirely on what you consider intense local work. Just because you are not compiling doesn’t mean running your 15 microservice stack locally is not intense.
Any memory intensive work. I thought his meaning was clear.
Memory intensive work wasn't the topic of conversation. GGP's comment was about how often he had to reboot his laptop despite it's immense memory budget, not about how much faster it was at performing memory-intensive tasks.
4 GB is maybe small nowadays, but memory-intensive is relative.
I'd argue that _most_ programmers or even data analysts don't need more than 8 GB to perform their actual work.
But realistically they need 16+ anyway because of opening 200 browser tabs, memory-heavy IDEs, etc. 32 or 64 is nice for sure, but definitely not needed for actual "work" workloads for the most part.
They need 16+ anyway because of running 2 or 3 Electron apps.
Having many browser tabs open (for research) is part of the 'actual work'.
>A colleague runs a Linux machine with 16GB of ram. Compiling the same project can cause OOM issues for him and the kernel starts whacking processes.
Which distro and can't you configure this? (Genuinely interested)
Also if the MacBook is better, than you should use that device. It's your "conclusion".
Of course it will run out of memory if it's taking more than 16GB of ram.
Most Linux distro suggest you to setup swap so instead of crashing it could use the swap. I personally don't use swap, because if something is consuming too much ram I most likely want it dead.
I don't know what are you compiling, but requiring more than 16GB of ram for compiling something sounds lazy and unneeded. Likewise you can use a browser extension which reclaims ram from unused tabs.
I can understand the need to have 64gb to do video editing or rendering, but most of the times we need that much ram is because our software is badly written and we don't want to face the problem, not because of the complexity of what we're doing.
> I personally don't use swap, because if something is consuming too much ram I most likely want it dead.
Swap is still useful to move stuff out to disk that _never_ gets accessed again.
Hahaha the old “if your workload just looked like my workload” argument.
I like how x86 is a power sucking hog for the performance and enthusiasts response is to go out and buy more power sucking hogs!
Someone wasn’t around when Linux could figure out what to do with a USB device. Spectre? Meltdown? What are those?
It’s almost as if engineering is hard and being a critic is easy.
Kind of like numbers, language allows for a never ending, so big in a sense, corpus of little ideas.
I never regretted moving from Linux to Mac also a decade ago. Mac has been fairly smooth for me, still on a 2013 mbp that just needed a battery replacement. Linux, however, was pain, quite often.
its a lot different now, unless your running it on proprietary hardware without official driver support, like the random stuff laptops have these days, but now we have many companies making laptops for Linux so it has improved a lot. Why give money to a company to use their software when you can use open software and help fix and improve something everyone can use for free.
Maybe not constantly but I have to restart my 64 GB Linux machine because it runs out of memory. Linux's OOM killer is terrible. I literally have to wait half an hour while my desktop is frozen for the OOM killer to kick in. And no, spamming the key combo for invoking the OOM killer once this happens doesn't help.
Use earlyoom or another oom daemon for a 1000x better experience and consider figuring out what is leaking. Do you use gnome? It used to have a bad leak that was supposed to have been fixed.
I think it's mostly firefox just holding things in memory since so much is free. Eventually it just all adds up and eventually I try to compile something and then it will run out of memory.
Yes, I can go setup an oom daemon, but Linux should handle this for me out of the box (ideally with some sort of logic to avoid killing something like X taking down every single program I have open). Getting stuck for half an hour or longer is just not acceptable to let happen.
Thanks I'll try that. The behavior when memory is low is easily the worst part of desktop Linux
I regret moving from macOS to Linux.
While the tooling and system stability are better, the desktop and app experiences are far worse. It’s buggy, half-baked, and crash.
I regularly hop distributions and do clean installs to see if that will fix it but they’re all the same — pop!_os might be slightly more stable in that regard.
I’ve been using Linux since 1992. I’ve had several year periods in the 90s and early 2000s where Linux was all I used.
I can ditch GNOME, KDE, and XFCE and go back to twm or i3 or fvwm or Window Maker, but I’ve moved beyond thoses experiences. I want a more functional desktop.
But, I see now why I paid more to put up with Apple.
Can you please explain what exactly you get from the Apple desktop you don't with any of those options? I'm currently forced onto a mac for work and I find it lacking in almost every way other than an ever so slight advantage in beauty.
So besides dpi-scaling issues linux is known for, whats the draw? Some hotkey that does a thing? Some visual candy? Smoothness in transitions?
Whatever it is, it's hard for me to imagine it being worth the tradeoff. I can't even name desktop workspaces on a Mac!
I'm reviewing the "new features" section on Monterey: https://www.apple.com/macos/monterey/features/ and I'm not seeing much other than a few known areas (multi-device support if you bought fully into the apple ecosystem, etc).
With most "I tried linux desktop" people it's a problem of DE/WM, but you seem to be beyond that so I'm genuinely curious.
> Can you please explain what exactly you get from the Apple desktop you don't with any of those options?
A more cohesive experience across the desktop and applications, eg.:
- The look and feel is uniform. Admittedly somewhat less so, these days now that the Apple Human Interface Guidelines are largely ignored.
- Mac apps and generally more stable.
- Mac apps generally have more features.
- The Mac apps I use generally tend to tie into the hardware (hw acceleration, drivers, etc) much better. The benefit of controlling the vertical, I suppose.
- The keyboard shortcuts both exist and are the same across apps.
- The design patterns.
- The AppleScript/x-callback-url, etc).
- Third party apps generally tend to fit well into the above.
On Linux, almost everybody seems to have their own ideas on how an app should be arranged. With full desktop environments, like GNOME and KDE lessen this, but it's still really, really common.
That said, my money is still on Linux and other open source OSes, because I don't want a single company telling me what I can and can't do.
yea like Mac not supporting Vulkan... or that Safari browser we have to make constant workarounds for, its the new IE5
MoltenVK is integrated into Vulkan now so Vulkan works fine on Mac
You mean Safari, that browser that does not kill your battery?
Not OP but I have extensive experience as a user of all three major desktop systems. Aqua, the macOS desktop, is by far the most stable and consistent, and in my experience the most usable and powerful as well.
A few things that stand out about it to me vs GNOME 3, KDE Plasma, Xfce, LXDE, and whatever the Windows interface is called:
- The consistency of the menu system, and being able to search and use the menus of any app from the keyboard with shift-cmd-?. This is like having Emacs M-x or Sublime/VSCode shift-cmd-P in every desktop app.
- Being able to assign custom keyboard shortcuts for any menu option in any app.
- Emacs keybindings for editing text in every text field (including on the web). I believe GNOME Tweaks is supposed to do this, but I could never get it to work reliably and universally like macOS.
- Native app ecosystem. Third-party Mac software is generally the most polished, though not quite as much exists as for Windows.
- System animations. This is a small one, but it makes things more fun and makes the whole system feel fluid and “organic,” for lack of a better term.
.
Linux works fine for me, but I don’t use a desktop. I totally agree with you that they are disasters, with poor documentation and multiple conflicting ways to make settings.
Firefox crashing is probably a Firefox problem.
You can set the scaling to be anything you want on X, just set the Xft.dpi number in your .Xresources file.
One problem with your Hackintosh install is that it’s closed source, no? As far as you know it’s exfiltrating your financial information to China.
>When people say Linux just works fine for them I genuinely think like am I crazy or cursed or something?
People who do that generally don't use Gnome, but i3/swaywm/xmonad and multitude of tools where you're in control what happens.
macOS still crashes, has weird bugs and every 6 months they go and change a bunch of shit that either deprecates functionality, causes conflict with how things work or takes another handful of your privacy.
Grass isn't always greener.
Edit for an example, I'm stuck on Catalina because they tried deprecating functionality my firewall uses. And the window shifter shortcut key program I use won't work in Big Sur.
> and every 6 months they go and change a bunch of shit that either deprecates functionality, causes conflict with how things work or takes another handful of your privacy.
No, that's Microsoft Windows. Apple only does it annually.
Windows has some of the best backwards compatibility around. Stuff from the 90s still works on Windows. Meanwhile macOS deprecated all 32 bit programs.
The kinds of things the post above complained about breaking on macOS—third-party software that messes with system internals in unsupported ways—also break frequently on Windows. To say nothing of the first-party breakage of Microsoft constantly moving around settings, re-enabling or re-installing annoyances you turned off, and removing your ability to easily get rid of them, and adding new ones behind your back.
Microsoft really only tries to preserve basic compatibility for well-behaved applications that aren't tightly integrated with any OS components or specific hardware/drivers and aren't doing anything that Microsoft disapproves of. Anything that strays outside those boundaries will run into trouble, and a lot of software ends up falling outside those boundaries even if it didn't really need to.
Both operating systems are pretty bad at letting you wield control over your own computer. But macOS doesn't try as hard to obscure the BS or obstruct your attempts to tame it, and macOS actually lets you refuse updates that you don't want.
Yes and many of the major Linux distros have discontinued or plan to discontinue official 32-bit/multilib support. Why? A lot of 32-bit libraries have serious vulnerabilities that no one is fixing. Essentially most of the world has transitioned to 64-bit except for a few key areas that are still trying to run libraries that are decades old.
Sure, macOS still crashes, I didn't claim it doesn't. But macOS crashes with a lot less regularity for me.
I use Linux as my daily driver and the apps I use (GNOME Terminal, other terms, Spotify, Slack, Brave and Firefox, and the GNOME desktop itself) crash multiple, sometimes several times a day on my two (work, personal) machines.
This happens across diverse machines, as well as across multiple distributions (though again, pop!_os, so far seems much less trashy for whatever reason).
what are you doing to cause these crashes or what hardware?? i use manjaro and pop_o, makes no sense since i get less crashes then windows and my computer is on for days without anything crashing, other then me writing a bug and crashing my own stuff :) i do lots of work and gaming on them so i don't know why your experience is so different.
Desktop Linux is easily the least-stable OS I use with any regularity. It was my main OS from something like '01-'10, but after I finally gave OS X a try, and since Windows got its shit together some time late in the WinXP service pack cycle (or, arguably, Win2k, but that wouldn't run my games) and stopped crashing all the time, it's really hard to justify using desktop Linux.
Does the OS hard lock or completely crash? No... unless you have graphics drivers issues, which isn't unlikely. Then, oh man, yes, lots. X/Wayland crashes that restart the window server? Yep. Applications crashing pretty regularly or glitching out so badly they have to be restarted, including the basic applications distributed with the heavier DEs? Yep. And it turns out that your windowing environment crashing or the main program you're currently using crashing is really close to as bad as the whole machine blue-screening in Windows, from the perspective of the user. Using Linux makes me anxious, even though I very much know WTF I'm doing with it. MacOS and even (spits) Windows don't do that to me, any more. Now that I've experienced not feeling that way, I can't go back. I go years between work-lost crashes of any sort at all on MacOS. Linux, one month without such a thing would be miraculous.
If you build up from almost nothing and keep things very minimal and have very boring and stable old hardware, and do as much as possible from the command line, it can be kinda OK, but it's a lot of work to set something up like that, and ongoing effort every time you do something manually that'd be automatic or trivial on a more full-featured GUI desktop. If you start with something like Ubuntu or Fedora standard desktop installs, though, there's just too much that can go wrong, and it will, with some frequency.
I could tolerate some hardware or workflows not working and things generally being a little less convenient, maybe, if it were rock solid, but it's very far from that. The main problems seem to be that its entire graphics stack is incredibly fragile (Wayland doesn't seem to have done much, if anything, to fix that) and it's way too easy for a glitchy driver to screw up the whole system.
I've switched to Linux from Mac OS about 2 years ago, being frustrated by memory consumption, slowdowns and freezes. After trying a couple of distributions / DEs, I've settled on Manjaro (Arch based) and KDE. Delighted with flexibility, features and general performance and stability. While, before, I was constantly tempering with the environment and changing stuff and there was always "something missing", I've found myself not needing to touch configuration or change anything in my workflow for more than a year now...I did have some glitches and CPU usage issues when playing YT videos on Intel-based graphics, but have since been using Lenovo laptops with Ryzen 4000 and 5000, and it's been flawless without any tempering. P14s Gen 2 AMD (5850U + 32GB RAM) is the best laptop I have ever used (software development), and I change them A LOT...
AMD graphics on Linux nowadays are vastly more stable than anything else. NVIDIA is proprietary crap and Intel Graphics still have cross-platform bugs so often is almost comical.
The experience you are describing is more akin to what I remember being Linux in the '00s than nowadays. I run a machine with arguably a buggy Ryzen 1st gen motherboard (it's quite buggy on Windows too), and I have a very stable desktop experience with Arch Linux and KDE Plasma. The real game changer has been AMDGPU, I'm yet to have any sort of graphical issue with an RX 580 on Linux, and Plasma is arguably quite solid nowadays.
The only issue I had recently were audio issues which were both due to my buggy soundcard and a bug in Pipewire. Excluding that, I've had almost no issues with desktop Linux since 2017.
I think your experience is not universal. I would say windows has been far less reliable than linux on my thinkpad machine, but overall I have to say that I have yet to use an OS that never crashed on me, all software of this complexity is buggy.
try viewing 2000 image files on Ubuntu desktop (20.04). It was so slow that I wrote a web interface to do the job instead.
Funny, I moved from Linux to Mac because I got so sick of no good laptops with full support for Linux and it's never ending subsystems for audio, webcams, trackpads, etc. And not like Linux UIs are known for being efficient and memory friendly either.
On the other hand my Linux computer struggles to go 10 days without a reboot due to the graphics driver bugging out, while my old MBA basically never needed rebooting.
This sounds appealing until you remember that you can't get all the nice apps in Linux.
Where did your comment come from, nothing in the parent comment is related to what you're saying.
Wasnt apple fined for slowing down iphones?
Looks like the 16gb model is now artificially made obsolete as well - you are supposed to buy the 64gb one.
For slowing down iPhones when the battery was low and not telling the user about it.
Stop trying to spin it Apple's illegal behavior. The change was clearly unnecessary because once they rolled it back the phones kept on functioning just fine.
Modern hardware can and should last ten years. Accept no excuses.
For slowing down older iPhones
I find the defense of Apple in this case absolutely fascinating.
You know how I can tell that Apple made that change not with the user in mind?
By the fact that until a 3rd party proved Apple was throttling older phones, Apple never publicly stated they were doing this. I’m fact, they vehemently denied it.
I saw this on a personal level, where my mother who took 2 trips abroad about a year apart, with the same phone (and she had a lot of disk space in both trips and she hardly installs any apps) found the phone an absolute pleasure to take pictures during the first trip, kept missing shots while waiting for the camera app to load in the second one.
And on taking the phone to the Genius Bar multiple times, she was gaslighted each time as the Genius showed her their diagnostic results which showed that no, her phone was in perfect health (minus a slight degradation in battery life).
Of course, a few months later the entire world learnt that, in fact, her phone had slowed down, and Apple was lying. What’s interesting is that Apple diagnostics tools (which should probably have a speed test?) also reported everything as being just fine.
For slowing down iPhones that would otherwise shut off due to degraded battery health, without notifying users.
It was a measure to increase product longevity (e.g a slow phone was seen as better than a phone that shut off randomly when voltage dropped)
It was also a measure that forced some users to upgrade.
My iPhone 5 got so slow and Apple was so unhelpful that I ended up replacing it. That was my first and last iPhone.
The battery that they made not replaceable. Interesting how it all fits together.
It is replaceable. It's just not replaceable without a screwdriver.
It is replaceable, and pretty cheap to get replaced at Apple too. I think it was like $29 or something to replace?
But even without Apple, that particular battery was pretty easy to replace
The $29 battery replacement program was only introduced after Apple was sued: https://www.macworld.com/article/230860/iphone-29-battery-re...
Also: users were not warned about this "feature", many (most?) didn't know why their phones slowed down, and there was no way to disable the behaviour.
Ah yeah, it was $50 before it was dropped to $29.
And I already mentioned in my initial comment that they didn't notify users. That's what the lawsuit in France was over, the lack of notice. I believe an OS update did eventually let users switch off the behaviour, and view their battery health.
why not just allow replacing batteries like phones did before?... seems like a dumb excuses to get people to buy a phone every year, make it impossible to replace the parts and slow down the phone. glue, solder, hardware lock parts etc etc until the phone is not even your property anymore.
the battery in that series of phones was replaceable quite easily, if you felt comfortable opening up the phone.
If the goal was to increase product longevity why did Apple only admit this wonderful measure only after it was proven by a 3rd party?
If this was such an altruistic measure why would the geniuses at the Genius Bar deny that my mothers phone had slowed down and throw diagnostic tests at our faces? Why was the exasperated Genius, who could clearly tell that the phone was slow indeed but his successful diagnostics meant there was nothing he can do, reduced to swiping up all the apps on the phone to kill them, when we both knew that if iOS was working properly that should do nothing?
Apple has few and long-lived product identifiers; as such it is possible for specific problems to become well known.
It's extraordinary to find more than a handful of anecdotes about any particular model number of consumer Windows laptop that happens to be on sale at Best Buy today. There are so many, and they cycle so rapidly.
thats the users every company wants!
I want to buy a 16" M1 Max because it looks awesome...but my base-model 13" M1 is a beast of a machine and I use it every day for dev work. It has only 8gb of ram but it still chews through everything. And I got it for a great price brand new considering how much I've been using it since launch. In my own experience so far, if I buy that expensive 16" M1 Max it will be because I want it.
Note that I've had Dell XPS laptops that were configured with the same price point as a Macbook and while nice...those devices definitely didn't have the same attention to quality that I've experienced with Apple products. So it seems unfair to make such a broad criticism against Apple users at large!
It’s not that amazing…
What is the alternative platform that runs Xcode but with the memory leak bugs fixed?
Have you tried not buying their shit?
Find me one laptop with a trackpad that works as well as a MacBook and we can talk.
I use a MacBook trackpad to make diagrams, for 3D modeling using Blender and FreeCAD, to draw schematics and route PCBs with KiCAD, and whatever else you’d otherwise do with a mouse, and it just doesn’t get in the way.
Over the years, I’ve been forced to use work supplied top-end Thinkpads, HP, and Dell laptops and none come even close to even the trackpad of my long retired, lowly 2009 MacBook Air 11”.
I don’t really care whether I use macOS, Linux or Windows: they all run the same applications. I also don’t care about the price of a laptop: what does it matter that a tool that I use day in day out for years costs $1000 more?
I care that it’s usable.
MacBooks are usable for things I use them for. The other laptops are not.
This right here, I got a Razer Blade from work last year, and found the touchpad so awful I assumed there must’ve been bad drivers installed. 3 driver installs later and I realized it just wasn’t as good as the MacBook trackpad, and I’d just need to pack a mouse everywhere I bring this laptop.
I love my 2017 mbp and I so want to tell every person posting workarounds for their $2000 brick please return it! Tim Apple won’t fix anything with all the $$$ in his eyes flowing in.
A finance professor once told me “if you don’t vote with your wallet you don’t get to complain”
Your mistake is thinking that that machine "doesn't work". It obviously works for them.
> Hundreds of millions of people, they must be crazy or very stupid! I must be right and they are the ones in the wrong!
That's how you sound.
So that's how people should react to a software bug?, move to a new OS and try to get as close as you can to all the apps you were using for years...
I think most everyone knows that this is a software bug that will be fixed.
Call Apple support as your experience is different than other people's. No memory problems here at all.
Apple support is generally not going to be able to help you solve software bugs.
I've had similar problems.. See my sibling comment, there's a chance it may be helpful!
Maybe try disabling TrueTone? Also, in Accessibility > Display try enabling Reduce transparency.
Next time it happens, can you try running heap(1) against it? It might help point out where that memory is going.
Just happened again [0] and no amount of app killing helps.
I ran heap(1) on WindowServer [1] but there's nothing in there standing out so maybe this is a red herring.
[0] https://cln.sh/A7YPLN
[1] https://gist.github.com/772687dba65a35cca8cd707fcfda99c0
Not a red herring, just confirming it wasn't a normal application memory leak (it's not, the malloc usage is fine–I think you're hitting the same issue as is mentioned in this article). The next step I would do is run sudo footprint $(pgrep WindowServer), optionally with -v to give you a really detailed list of why the footprint is so high.
Off topic but I love Lunar. Thanks!
Being serious: Do you have Chrome installed? I have been following this story for a while and nothing seems to be coming out of it:
https://chromeisbad.com/
https://bugs.chromium.org/p/chromium/issues/detail?id=115840...
I do not use Chrome and my windowserver is using 3GB after a month of uptime. I feel this is not a M1/Apple issue to be honest.
> sluggish performance of a 2015 iMac since practically the day we bought it
I wish I had found this website when my MacBook Pro was still alive, maybe Chrome actually killed it after all. The low performance was so bad I moved to Windows 10 ("the last Windows system" they said), but after it notified my system can be upgraded to 11 I quickly moved to Ubuntu. I'm currently experiencing a few hiccups and reduced them with some memory tweaks, but still the experience is much better than Mac and Windows.
I switch over to safari for the time being and no issues.
I feel like I'm a bizarro world, because I don't get this stuff: crashes, memory leaks, and so on.
I use: VScode, Idea, iTerm2 (with x86 brew), Logic (and lots of plugins), FCPX, Live (x86, and lots of plugins), Slack, Reeder, MS Office, Spotify, Chrome and other stuff. MBP M1 16GB.
What I do get, is some bluetooth issues (e.g. affecting the mouse cursor. If you start the bluetooth deadmon it fixes itself for a day or so.
You're not in a bizarro world. Most users are pretty happy with their M1 Macs (myself included) and HN just tends to upvote anything anti-apple even if it's an isolated experience.
I'd wager most people on this thread do not actually own an m1 Mac.
My experience (and that of others I know that have them) has been nothing but stellar.
HN has more Apple "fans" than "haters". And they will downvote anything bad about Apple lol
I own a M1 8 gb and It has to go through multiple reboots/day with lot of open tabs/apps on Monterey
Was it doing that before Monterey? I haven't upgraded to Monterey yet because I want to avoid any bugs that haven't been ironed out yet...so far my M1 8gb has been excellent and if those issues are introduced by Monterey then I may just delay upgraading until next year sometime.
Not sure because I quickly upgraded it and I haven't a lot of open pages/process before
Worth a read if you ever wonder why certain people on here are oddly fixated on the platforms other people are using. (Especially when they have no clue what you even do with your hardware.)
https://www.theatlantic.com/family/archive/2021/11/group-nar...
I’m generally happy with my Mac products, but I’ve also gone through long periods where my Mac would crash multiple times per week due to serious OS bugs.
Most recently, the Thunderbolt 3 issues were responsible for countless hours of lost time rebooting my laptop last year, and I’m not alone: https://rachelbythebay.com/w/2020/10/03/repro/
The weirdest part was that this occurred with only Apple products connected to my laptop. I couldn’t even blame it on 3rd-party products. It also worked perfectly fine before and after the range of affected MacOS versions. There were many reports of the same issue all over the internet, so I know I wasn’t alone.
Then one day Apple finally fixed the bug and everything was back to normal. But all those months of constant reboots and crashed and failed workarounds left a mark on my overall experience.
So while the Mac experience is generally good, it’s still far from perfect. Weirdly, I’ve had far fewer problems with my Windows system in recent years.
Just a wild guess, but I imagine most users are not pro users and don't even notice if a process leaks memory.
My Mac used to expose the same kind of problems, that's why I switched to Linux on non Apple laptops years ago.
My feeling was that the quality of the OS went down constantly and apparently it's not coming back with newer releases.
Your position is that HN users are anti-apple and lying about this issue, rather than people with no agenda just sharing their experiences?
Claiming most people don't have M1 apple products is also bizarre considering almost all startups are using exclusively macs and we've all been waiting for the this release.
Its not like Apple doesn't release faulty hardware and software. Butterfly keyboard is a good example of faulty hardware and OSX Lion crashed all the time on the Macbook pros with discrete GPUs till Apple patched it.
I don't think it's really anti-Apple, but these articles do tend to take what seem likely to be very isolated events and present them as widespread, systemic issues.
I have absolutely no doubt that this person is having bad memory leaks (though their narrative about SoC memory or other elements seem unlikely to play any part whatsoever). Software has faults, and something in their environment or stack is causing problems. I have an M1 Mac and used every beta of Monterey, and now the release, and have had zero problems. Like nothing at all. I use XCode, IntelliJ, Chrome, Safari, Excel, Logic Pro among others all day long, heavily. Loads of other people seem to be having no problem with their systems. Eh.
If something has literally millions of users, saying "{X} happens" will invariably draw out someone with a similar anecdote, but that doesn't mean it happens to everyone, or even a significant minority. Some unique combination of factors is yielding a poor user experience, and hopefully there is a resolution, but it just doesn't seem likely that it's widespread.
elephant in the room for all these sort of articles it the law of large numbers. Sell a couple million devices with a defect that affects .1% of people and it'll be a thing for 1,000s of folk. And apple sells 10s of millions.
Unfortunately it’s comments and religious defenders like this that allow Apple to get away with selling broken products for years, and then issue recalls for products sold 4-5 years ago when people are not using them anymore, like they’ve done with multiple Mac laptops that had graphics card and screen issues.
Same here. My 16-inch M1 has been a dream so far.
Yep. I bought my M1 13" (with 8gb of ram) when it launched and it's easily the best laptop I've purchased. I want the 16" with the M1 Max because it looks incredible but so far the M1 13" has chewed through everything I've thrown at it, and I use it heavily every single day. Fantastic laptop.
> HN just tends to upvote anything anti-apple even if it's an isolated experience
Why do I see so many pro-Apple posts then? And even posts directly pointing to an Apple ad.
Im willing to bet this community has a higher percentage of apple users than most others. The number of people claiming to already have an M1 of any kind is surprising, especially considering Ive received a macbook pro for every single software engineering job Ive ever had.
I think the main reason why these stories are brought up for Apple so much is because it has this persistent "things just work" narrative that a lot of people buy into when they decide to pay the Apple premium. When it turns out to not be true, it stings more than similar issues on e.g. Windows or Android, simply because the expectations there aren't that rosy to begin with.
Please don't make spurious generalizations about the community—these images are nearly all in the eye of the beholder, i.e. people with the opposite preferences to yours see the community exactly the opposite way (and make similarly spurious generalizations about HN favoring Apple or whatever $BigCo they don't personally care for).
Because such generalizations are just encodings of personal likes/dislikes in the form of general claims, they lead to lame discussion.
https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...
https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...
https://news.ycombinator.com/newsguidelines.html
I have had really bad issues with bluetooth mice on my M1 macbook, the lag is so bad. Bluetooth is basically unusable for me so I had to switch to the logitech unifying receiver which has been flawless. This seems to be related to using airpods at the same time.
Otherwise I have been very happy with the laptop.
I love that Apple hasn’t been able to fix their Bluetooth bugs in forever. I always hope they’ll be fixed with the next OS version, maybe the next MacBook, maybe the next iPhone.
Nope. My AirPods have regular issues and my Magic Mouse regularly disconnects.
I hoped my next M1 would solve it. Guess not.
I don't know that it's a Mac thing - I get similar Bluetooth issues on Windows. Honestly, Bluetooth just isn't that good.
The best software stack for bluetooth audio is the one provided by Pipewire on Linux.
Pipewire for some reason insists on using GTK4 for its UI though, which hangs X cold if you use any of the popovers in it. I didn't even know it was possible to hang X like that.
Is that an X issue or a GTK issue? Haven't heard of this before and it sounds scary
What do you mean by a Pipewire UI? It's a background daemon (or three: pipewire, a session manager, and pipewire-pulse).
EDIT: Perhaps the parent means Helvum, which ldd says links to libgtk-4.so.1 => /usr/lib/libgtk-4.so.1. However, I can't figure out how to open a popover (https://python-gtk-3-tutorial.readthedocs.io/en/latest/popov...) in the app; all I can do is drag to create or break connections.
Try to set up https://github.com/jaakkopasanen/AutoEq impulse responses for headphones, and you _will_ inevitably hang your X. Daemon isn't terribly useful on its own if you can't set it up to do what you need. So anyone considering Pipewire as a "better" alternative to anything should also know, at a minimum, that it's difficult/impossible to uninstall and buggy AF.
To clarify things, PipeWire is an audio daemon, completely unaffiliated with https://github.com/jaakkopasanen/AutoEq which is a set of headphone impulse responses. It just so happens that the recommended way to apply these impulse responses on Linux is EasyEffects, which now depends on both PipeWire and GTK4. Nonetheless, PipeWire is not EasyEffects and not AutoEq, and "Pipewire for some reason insists on using GTK4 for its UI though" is untrue because EasyEffects is not PipeWire's official GUI, and doesn't even come up when I search `pipewire gui`.
I haven't tested EasyEffects. With that in mind, PipeWire is buggy AF for the time being, and I've been working over the past week to clean things up (fixed a few bugs in plasmashell and pipewire so far, but I'm still learning and there's a great deal of bugs I haven't diagnosed and explained to the maintainers, it will probably take weeks longer for me to work through everything, and I'd very much prefer the program was written to handle race conditions and edge cases properly, instead of leaving users to discover bugs and trace the root cause through the labyrinthine IPC-heavy event-driven codebase after the fact).
Their choice to write new code in C at this late date is hard to justify. A bit of abstraction and automation can go a long way.
My guess is that Pipewire is written in C because the author likes C (or at least somewhat dislikes C++). And it has "abstraction"... in the form of hand-baked vtables and observers created out of nested macros, void*, and intrusive linked lists.
An advantage of C is that it comes with a stable ABI, where Rust is unstable and C++ is a subject of conflict (though you can achieve a stable ABI by restricting your external API to C functions, and if you want cross-shared-object virtual calls like Pipewire, you might have to write the same pile of macros as PipeWire, which you were trying to save yourself from). The downsides are memory and type unsafety (surprisingly the only crash I've seen so far is a null pointer dereference, and I haven't hit any obvious memory or type errors), and a complete absence of "vocabulary" or templated types (they've created their own slotmap-like ID allocator but without a generation count resulting in ever-present "logical use-after-frees" in library and application code, and an intrusive linked list which they use even when hashmaps might provide somewhat faster iteration and asymptotically faster search).
I assumed the C ABI, and thus the cobbled-up vtables. Where the author might have used C++ to advantage would be automating other, internal details of the code. There is nothing quite like bugs you don't have because you didn't need to write the code where they would have been, or because you could run code at compile time to generate or at least verify your cobbled-together apparatus.
The saving grace is that it can at any time be switched over to be built with a C++ compiler, and then start to be modernized incrementally.
What you say is true. Unfortunately I don't think you can trivially switch it to a C++ compiler, because they use {[FOO] = ...} initialization.
It would be cool (not necessarily useful, probably not a good use of time) to rewrite it in Rust while keeping the dependency count in the single digits. It would be helpful to contributors to add architectural documentation.
Clang supports it in C++ mode but not GCC sadly :(
C++ has a version of that notation, with minor (and IMHO unnecesssary) restrictions that probably impose no hardship.
C++20 added dot initialization, which has been supported in MSVC/GCC/Clang for over a year now. No released version of the C++ standard supports array index initialization, and jcelerier says Clang but not GCC supports it as an extension.
I use qjackctl as "pipewire UI", works fine
Does it handle headsets correctly?
Asking the real questions here.
I finally gave up on using an XPS15 with Ubuntu because I could not get over the fact it cannot do high quality bi-direction bluetooth audio.
Edit: Looks like it does offer some improvments over pulseaudio: https://www.redpill-linpro.com/techblog/2021/05/31/better-bl....
I finally figured out how to get bluetooth headsets to work correctly on my desktop: output to 3.5mm, 3.5mm to dedicated bluetooth transmitter.
Throw hardware at the software problem, because software absolutely cannot be trusted not to foul this up. Not Windows, not linux, not pipewire, not pulse, none of it. Evidently not mac either, though I've personally never had trouble with bluetooth on mac. Bluetooth is 23 years old and its flagship application is still completely broken across huge swaths of the client/server compatibility matrix.
I switched after seeing PRs improving Pulseaudio get shot down pretty dismissively, and iirc the would-be contributor getting fed up, abandoning it because well whatever, Pipewire.
But actually, since switching the AV sync is off using my Bluetooth earphones, at least in Firefox.
Curiously, I can fix it with latency adjustment in pavucontrol, but it only has an affect when the 'advanced' drop down is revealed in that tab of the menu. Close it, and it's off again. Open, and the value has persisted, and it works again. Haven't really dug into that because I've been hoping the (presumably) underlying Pipewire issue will be fixed.
Oof, I saw that PR drama too (~half of bluetooth headsets were broken, PR fixed it, got ignored for months despite loads of 3rd party attention, author got heated, maintainers show up very interested in swinging their CoC around but very uninterested in merging the fix, author gave up and left for pipewire).
Pulse is still broken a year later, pipewire is fixed.
So it goes.
For the curious: https://gitlab.freedesktop.org/pulseaudio/pulseaudio/-/merge...
Pipewire is not written by or maintained by the same people who did PulseAudio. Thus, such experience likely does not carry over.
That is not to say Pipewire could not also be a problem, but I haven't encountered such complaints about it.
> Honestly, Bluetooth just isn't that good.
I think you're right about this. I have a pair of JBL headphones and it will just randomly become completely garbled during a video conference under Windows.
I have to disable/re-enable Bluetooth to "fix" it. It's a mess.
Conversely, I have a couple of Bose QC35, and either one of them works with pretty much any device I tried to pair them with - Windows laptops, Android phones and tablets, and a Chromebook. So perhaps it's a QoI issue?
I've had similar issues with Bose QC35 on Windows. Currently they work -- but the settings are on a knife-edge: all it takes is an unlucky Windows update and I'll spend half a day trying to find the right services and settings to reset/reinstall to get it paired again. (Also, whenever I restart, I have to open the bluetooth troubleshooter to get them paired, though at least that always works.) I blame both Windows and Bose equally, but I also think it's a very difficult protocol to work with.
I've had an issue with Windows 10 cutting off the first 0.3-0.8s of any audio that plays over Bluetooth for a while now. It's occurred across entirely different systems with entirely different bluetooth chipsets and entirely different headphones. Nothing is a surefire fix except for keeping a silent audio file playing in the background, which is ridiculous.
The bug isn't that bad if all you use bluetooth headphones for is YouTube, video chat, and music but it makes Anki language learning decks with audio almost unusable.
I think it's not that it isn't "good", it's more that it is an enormously overcomplicated design-by-committee. The spec is really complicated and downright weird in a lot of places. Kind of like they were trying out new academic ideas that weren't really practical.
I'm more familiar with BLE, which is better, but even that has a completely weird "characteristics" protocol that you have to use. Want to send a big blob of data? You have to manually cut it up into 20 byte chunks to send it over "notifications".
It's like they looked at the OSI model and thought "ah yes this looks right! we'll implement all the layers!"
Anything below Bluetooth 5 is barely usable. The 2.4Ghz freq gets interfered by anything from fluorescent lamps, old Wi-Fi hotspots, old laptops and phones to microwave ovens, and if it's not yours, it's somebody else's. You basically have no control over whether your bluetooth devices get interfered.
I used to have similar problems. It turned out to be Bluetooth congestion. When I relocated to an office away from a window that faced a public street, the problems went away.
Apple is still using Broadcom chips for BT, which might be part of the problem.
Doesn't really make sense for a new CPU/GPU to fix the problem, what you want is for Apple to stick in their own networking cards.
Airpods Max are difficult to connect to macbook pros
Get your iPhone away or off can help. It seems they always want to go to the phone first.
The cursor thing is frustrating. When it lags it makes me suspect someone has a hidden instance of VNC open watching my screen. Perhaps they do.
When porting our audio plug-ins to Big Sur we had to avoid some APIs that were running fine for years, because of new memory leaks.
Both in this article and other comments here, people mention Safari being a factor. It sounds like this is likely related to the browser. I noticed that you're using Chrome. I use a variety of browsers but rarely Safari, and I don't have any of these issues either.
It is always Safari.
I saw in a screenshot in this thread Safari with 15GB or so for a single page (and I think it was just some newsite or such)?
That sounds like a big bug and a very possible cause.
I find all three (Linux, MacOS and Windows) to be fairly stable. It's been many many years since I've run into either BSOD on Windows or the beach ball of death on Mac.
My guess is faulty hardware. I've had plenty of those on the 2018 MBP (keyboard getting stuck, bulging battery and dead zones on touchbar). Waiting to see if the QC is better with new M1 Pros before buying.
>What I do get, is some bluetooth issues (e.g. affecting the mouse cursor. If you start the bluetooth deadmon it fixes itself for a day or so.
Can you elaborate on this? I think i was hitting this as well. Specifically when I scroll or type, it "halts" for a split second every other second. Originally I thought it was a video card failing or something but after a long time I finally traced it to Bluetooth. Since then I have been using wired connection on my keyboard and mouse.
What do you mean by "If you start the bluetooth deadmon"
In my case it's visible in the mouse (the keyboard is wired, mechanical), and it's like you describe, at certain points there is laginess. After some time, it's like the mouse slows down, and lags more.
I've heard this happens when you have some USB-C hubs attached, which I do. Not sure what the connection is -- but I think my Anker hub does causes this.
>What do you mean by "If you start the bluetooth deadmon"
Sorry, I meant: "if you restart the bluetooth daemon", e.g. "sudo pkill bluetoothd".
See my response to you above...
a little unclear if you don't have a hub, or do have an Anker hub, but try moving the Anker hub far away from your Macbook and see if you still have the bluetooth issue. You need a USB cable (depending on what's the hub is you might need a 3.1 Gen2) long enough.
For me it was a hardware proximity issue. Powered external devices interfering with the Mac's bluetooth.
Same here - I've been slamming my 13" M1 base model (8gb ram, 512 gb storage) and it's a beast of a machine. I throw a lot at it for work every day - multiple browsers open, back-end processes, the UI app, pgAdmin and postgres both running locally...it crushes everything I throw at it, with great battery life too.
Only issue I've seen with it is some bluetooth flakiness sometimes the way you mentioned. Other than that, it's been a rock solid device and the only reason I haven't upgraded to one of the new MBPs despite really wanting one...is because so far I haven't been able to justify it yet considering how rock solid the M1 has been!
i have an older macbook pro (only for another week until my new one arrives), and i found switching from iterm2 to kitty improved performance of my computer drastically. i think i have some other HD corruption issues going on, but kitty extended the lifespan of my computer such that i was able to wait for the new pros.
When I get Bluetooth issues, I just power cycle my trackpad and keyboard and that seems to do the trick
My Bluetooth issues are clearly hardware related. Out of all machines I have it only on my iMac. Software is almost the same on all Macs.
This whole thread is striking a nerve with me since these seem to be all of the issues i had been dealing with since getting a new 16 inch a few weeks back. Specifically "windowserver" and also these bluetooth issues.
See my other comment in this thread about how i solved "windowserver".
Bluetooth issue was a hw proximity issue for me. My CalDigit hub was mounted right below my Macbook. I noticed when I moved my MacBook's location (still connected to hub), the bluetooth issue when away.
It's worth a try if you have a powered accessory near your macbook and haven't tried this already, but...
I do still get the issue once in a while, but I can now at least just turn the keyboard/touchpad off/on and it goes back to normal whereas before, it happened all the time.
Once in a while I still have to Shift+Option+"bluetooth menu icon" and reset bluetooth module.
I know these are just work arounds, but at least it allows me to get back to work and only have to deal with it weekly instead of not being able to use bluetooth at all.
Edit: just upgraded to 12.0.1 and noticed Shift+Option+"bluetooth menu icon" to reset bluetooth module is gone!?
It’s probably USB 3.0 frequencies clashing with Bluetooth freqs. Even worse if you’re connected to 2.4Ghz wifi since it’s the same frequency Bluetooth uses.
I have been using an M1 MBA for nearly a year, and I have never faced any issues or memory leaks. It has been solid as a rock.
I recently had to use a Windows laptop for work, and god I hate when windows becomes "not responding"
If windows is not responding error it's an app bug and not a windows bug. I have to swap between windows and mac all the time and both are perfectly usable when you have enough memory for your given task, and that is 95% an app issue as too many modern apps drink memory like they just came out of the Sahara.
I've had explorer.exe crash many many times.
Explorer hosts a million plugins for everything that wants to do things like add a menu entry to the desktop right click menu, and each of them can crash or hang Explorer.
It’s just a terrible design.
Wait until you hear how Finder handles things, I honestly couldn't believe my ears the first time I asked "why can't I close the filepicker?"
Can you explain?
There’s similar interfaces for Finder for things like Dropbox, except not every mouse driver has a plug-in and if it turns out the model is problematic, like on Windows, Apple has no issues with repairing it and forcing the plugins to update.
Backwards compatibility is nice but it drags you down.
Explorer is pretty stable if you don't have any shell add-ins loaded.
Since we all seem to be sharing that we don't have issues with our OSs, I've never had much issue with any NT based Windows systems. I've had bad hardware that once replaced improved reliability, but generally speaking I can go months without reboots. (except for those corporate forced updates for security reasons, but those exist on every platform, so I don't consider that an issue)
https://www.youtube.com/watch?v=ReHafyiDTR0
On the last few versions of Windows, it doesn't really feel like my computer anymore. I use my OS to lay out the structure of my thoughts, ie. I'll have text documents open, arranged in a certain way, project folders open, image editors with sketches in them... then Windows Update wakes my computer while I am sleeping and force-closes everything; I awake to a blank slate. It's as though someone has snuck into my house in in the middle of the night and swept everything off my desk and into the trash.
Block the windows update domain names or get an application firewall (fortinet makes some cheap ones) and disallow windows update.
It does suck that you gotta do this, but at least there is a way.
I'm glad new editors auto save everything and can restart without issues. Just a shame the os doesn't do it for all apps.
I’ve had a similar experience with an M1 Mac Mini. I’ve thrown the kitchen sink at it and it doesn’t break a sweat.
Anecdotally, it feels like I must restart WindowServer in OSX more often I've had to restart explorer/dwm in Windows. But neither are particularly reliable, in my experience.
Do you by any chance have the base model? I'm trying to see if folks think it's sufficient for their needs. It's hard to calibrate memory needs coming from Intel.
I have the base model 13" M1 (8gb ram, 512 gb storage) and it's rock solid and sufficient for my needs. I use it every day for dev work. Got my UI app running on it for local development, backend processes, connections to Heroku and my local postgres db, multiple browsers open, music streaming, multiple bluetooth devices connected to it...
Performs like a champ with excellent battery life. I keep waiting for it to start giving me some slowdown as our apps become more complex, so I can upgrade to one of the shiny new laptops...but so far, it's been a champ.
Are you worried about swap use at all? I’m on the same boat but I’m a noob so I just put my IDE and browser tabs on my memory. I don’t even know how to use docker yet
Not at all. I haven’t had issues with swap use so far and I intend to upgrade on a regular cadence (every 18-24 months) so I don’t expect it to be an issue for drive longevity.
It's probably worth reminding everyone who is new to the Mac or like most of us keep forgetting. Don't update your OS until at least half way through its update lifecycle e.g. for Monterey wait until 12.0.5.
This happens every single year where there are annoying/serious bugs that takes a while to get fixed as the focus is on P1 showstoppers.
I tend to stay 1-2 years behind. But that hasn’t been an option on new hardware since at least the early OS X days, if not maybe Classic MacOS.
Yeah, I recently updated to Catalina. I had to Google to find the download link, but it was still available from apple.com
Yup. I upgraded from High Sierra around the time Big Sur rolled out.
To be honest, I’m probably not going to upgrade again at all until the new MBPs seem stable enough for me to buy. Nothing in Big Sur or Monterey seems compelling enough, apart from M1 support and access to iOS apps, for me to risk the disruption.
You can download the full installers from the command-line: https://512pixels.net/2019/10/download-macos-installers-with...
Wikipedia shows the exact version number of the latest releases for each.
My primary is still in Mojave and I’ve been perfectly happy with it.
Now, if I were to update my hardware that won’t be an option.
Still an option. I updated to 10.15 "Catalina" (released June 2019) about a year ago, and am still on it.
I will probably soon-ish update to MacOS 11 "Big Sur" (released November 2020), but not yet the just-released MacOS 12 "Monterey".
> Still an option. I updated to 10.15 "Catalina" (released June 2019) about a year ago, and am still on it.
On new hardware released with Big Sur? To be clear, what I was saying is not an option is “downgrading” new hardware to an older major version than the one it was released with.
oh ok, I misunderstood!
Hah well with the M1 air I got a few days ago, it helpfully won’t let me upgrade to Monterey! I’ve downloaded the 12GB update three times now and it fails each time it tries. I guess I’ll wait.
I was also having trouble installing final cut/logic. I ended up needing to delete a few AppStore cache folders, immediately resolved the issues.
click through the various letters and delete any com.apple.appstore folder contents. reboot
I do that as well - I have work Mac and personal one, and I don't want to run them with different OS versions since Apple loves to change things between versions for no reason (last time they renamed "grab" to "screenshot" and needlessly reordered System Preferences) so using two versions of Mac OS is super annoying. And corporate takes a while to get all their crap running on a new version.
12.1 is already in beta.
Currently 12.1 has this issue though.
I saw this recently primarily with Firefox and (presumably?) Firefox may have been updated to fix the issue because it is thankfully not happening anymore.
When it did happen though, it seemed ludicrous: that a single app could in a matter of literally minutes start burning through gigabytes of memory, to the point where the entire system seems to freeze with no option except to kill Firefox (and everything it may be doing) to continue. And, like the author, I had started keeping Activity Monitor open so I could pre-kill Firefox at a point when I knew it would probably start to be pushing it.
It never feels good to baby-sit a machine. It feels even worse when it’s an expensive new machine. And it means that the operating system is failing at its most basic task (arguably the entire point of having an operating system) — to manage resources well.
Likely this OS bug regarding mouse cursors. Improved but not fixed in Firefox 94:
https://bugzilla.mozilla.org/show_bug.cgi?id=1735345
> And it means that the operating system is failing at its most basic task (arguably the entire point of having an operating system) — to manage resources well.
Before about 2001, that would be a bold claim indeed in the Mac world. Pre-emptive multitasking only became a thing with the transition to BSD-based OS X
Sounds very similar to the macOS 10.15.6 release, which had a horrible kernel memory leak. It was especially visible when running virtual machines. After a few hours the kernel memory pressure was so high that it would oom kill every usermode process one by one, including critical system processes.
That one had me yearning for a linux grub menu entry for "previous kernel".
Is kernel memory pressure different to the normal memory pressure? Or is it the same thing?
It's fundamentally the same thing, but when the kernel is refuses to deallocate memory you can't just terminate the process, you have to reboot the system.
The kernel memory usage would only show up with ever-increasing numbers in "sudo zprint", it wasn't visible in activity monitor
MacOS has had intermittent memory leaks for quite some time. Even my 2013 MBP running an old OS version has all the Apple processes slowly leaking memory until I reboot.
I doubt it's directly hardware related, though it's possible it is scoped to special MacOS logic that only is executed on M1. My assumption has been that it's actually coming from shared libraries that Apple-owned processes share, because I usually don't notice it happening for non-Apple processes.
Which OS version? Personally I'm on 2015 MBP with High Sierra 10.13.6 and zero memory leakage. I go months without even turning off the MBP. Granted, sometimes I get memory leakage behavior from Chrome/Firefox. I usually stick to Safari.
No problems for me with my M1 Mac. Heavy use of compilers, numerical tools (both native apple silicon and some that still need to run under Rosetta), a couple different editors, some specialized analysis tools, and the usual set of desktop apps for day to day work and fun. My workload tends to be pretty demanding of a machine (modeling+simulation where I can soak up every core and bit of RAM available to me). I can’t remember when I had to reboot the machine other than OS updates.
For people who are experiencing a leak on macOS 12: Check if you have set a custom cursor color in the system accessibility preferences. Resetting the cursor color to the default may fix it.
See https://bugzilla.mozilla.org/show_bug.cgi?id=1735345#c18
(edit: rephrased to mention macOS 12)
Good lord would I love to see the postmortem breakdown on that one
Having worked at Apple adjacent to the likely responsible teams, the relevant details are: overworked engineering teams, insufficient QA investment, and most of all a yearly macOS major release schedule that persists primarily to benefit the egos of Apple's upper management.
Until you see a WWDC that doesn't announce a major release of macOS, expect more of the same.
To any involved Apple folks here, I feel your pain, and I'm sorry.
And this is why I have completely stopped using all Apple products. I used to be a massive Mac fanboy. But the software quality is absolute shit now and has been for at least the past 5 years. I don't understand why anyone puts up with it.
What do you use now?
I speak only for myself, but after Catalina I started transitioning to Linux and couldn't be happier. Being able to set up my entire computer with a single Github repo is a godsend for productivity, all it takes are a few keystrokes and I've got a fully configured desktop with all my dotfiles, keybinds and applications.
I've been a Linux user for some 10 years, but have also had to use a Macbook Pro for a while due to work. There's nothing stopping you from fully configuring macOS through a single repository similar to how you do Linux. For a while my dot files were aimed at Linux (primarily), macOS, and Windows as I anticipated to have to use both macOS and Windows down the line, though I've since removed much of the multiplatform code.
Things may be different now as I haven't used a mac since about 2015, but depending on what tools you needed you couldn't escape the GUI completely (for a fully automated setup). At this point I don't remember all that was wrong, but one of the biggest required installing xcode or something like that before I could get git to work.
Things have been different for many years.
a) You can install XCode CLI with xcode-select --install.
b) Many applications can be installed with homebrew casks.
e.g. brew install --cask docker google-drive intellij-idea microsoft-office
You don't even need --cask anymore. You can also use `brew bundle dump` to synchronize what you've installed via brew, and it also can sync mac app store installs too.
It's a bit annoying getting macOS system preferences synced still although.
How are you managing the desktop (gnome) files? I swear every time I try to move them to a new system or account all I get is grief; weird bugs, application crashes, dbus weirdness. I would love to see your recipe :) . All I can usually do is bring over app config files like emacs/vim/etc/personal scripts.
I don't use GNOME on any of my systems anymore (the 40 update destroyed all hope I had for the desktop going forwards), but I'm convinced you can get some settings transferred by poking around in your ~/.local directory. Having written a few GTK applications in the past, it seems like the rationale for this is because of GNOME's settings API, and how they define compliance with the GNOME spec. I find it to be a total clusterfuck, and if I had any confidence in the desktop's current management team I'd probably open a few pull requests/RFCs to try and fix it. I might sound a little heavy-handed here, but the "my way or the highway" rhetoric that GNOME's developers are pushing right now makes it really hard for me to take their desktop seriously, or blame anything but the developers themselves for their lack of features.
In any case, I can assure you that behavior is perfectly "normal" to them, unfortunately.
All good points. I haven't really tried with kde. Thanks!
Recently I discovered Extension Sync[0] for GNOME, I never really used it, but it may be useful to you.
[0] https://extensions.gnome.org/extension/1486/extensions-sync/
Thanks, I'll check it out.
That sounds really neat is it some thing you can share? I recently setup a new machine which didn't take very long, but I have run into a few things I forget so would like to set up something similar.
I basically just run an install script which does a few different things. I start by updating the repos and regenerating my mirrors so I get the fastest downloads possible. After that, I install shell utilities (my editor, shell of choice, base-devel, etc.) and then I enable the AUR so I can grab the rest of my desired apps. The rest of it is decidedly basic, it just copies my tracked config files to ~/.config/ and moves my wallpaper to /usr/share/wallpapers. I run a few rain dances automatically too, like `sudo chmod a+wr /opt/spotify` (which lets me automate my Spotify theming process) and installing/unzipping a Steam theme.
To install, I run `git clone https:github.com/username/repo && ./repo/install.sh`, and I'm off to the races. I really reccommend writing one for yourself, as it's a great way to learn shell scripts.
Thanks for that.
For me there's a trade-off, depending on how often I setup a new machine. Value of automating with a procedural script vs taking the opportunity to try a leaner (or more modern) toolchain.
Currently upgrade about every 2-3 years, so I do tend not to need a bunch of things each iteration, and might upgrade OS. But I'd like a backup in case e.g. hardware failure, laptop lost or stolen.
Some declarative workstation config like terraform or Ansible would be interesting.
Care to explain what you are talking about?
This isn't exclusive to Linux.
With homebrew you can install most applications through the command line as well.
> Being able to set up my entire computer with a single Github repo is a godsend for productivity,
Not sure I understand, could you elaborate what you mean.
Android phone and Ubuntu on my desktop PC and ThinkPads.
I've been seeing this exact same comment for over a decade now.
The first couple of releases of OSX are always a little rough around the edges. And then by the end of the cycle everyone proclaims how wonderful this release is.
See everyone next year !
Except the bugs have continued to get worse and worse. For me the deal breaker was the blank page in iOS Safari[1]. After fighting that one constantly, I threw in the towel. I've been so much happier. Btw I was an Apple user from System 7 through to 2019.
[1] https://forums.macrumors.com/threads/safari-blank-page-bug.2...
> worse and worse
https://www.zdnet.com/article/users-complain-of-mac-os-x-lio...
Back in the days OS X was a dumpster fire in terms of overall reliability until the very late releases of every major version. Some versions like Lion had pure crashes until the end.
Even looking a few years before, we had the root user password bug and other pretty severe issues way worse than what we’re seeing now.
I did say over the past 5 years, so things like the root password issue I include in that. I'm also utterly convinced airdrop is just a long running joke at Apple, I've gotten it to work like twice.
Snow Leopard was as solid as OSX ever got in my estimation. Man I miss those days. Snow Leopard's primary goal was just to harden Leopard and not really have too many new features. Apple really needs to bring that kind of thinking back. I'm happy to see their hardware is back on track, now if they could just get the software to do a U turn...
Craig Ferengi has to go. Too much of a pretty boy who doesn’t get the need to invest in QA.
I have always felt Macs are awesome hardware bundled with poor OS. Apple needs to up their software quality.
What I don't get about apple, with a compensation policy that is less than the other big techs, because they can't get over the late 90s, team sizes that are usually smaller and being one of the largest market cap companies in the world, why don't they hire more?
Hiring more people isn't going to make them more money. Improving their products/services isn't going to make them more money. Providing more value to customers isn't going to make them more money.
Charging $300 for a screen replacement will make them more money. Selling bluetooth headphones with non-replaceable batteries will make them more money. Collecting taxes on developers will make them more money. etc etc
They got popular because of their commitment to quality. "It just works" is and in many ways still is a key to their popularity. They whittle away at their "quality credit rating" the less and less they invest in these things, until their quality credit rating becomes low enough for a competitor exploit that and make them lose market share very quickly. It does make them money.
I can see the business side of the permanent upgrade hype, constant new features, etc. But aren't we supposed to be reaching some kind of detente with Apple where they realize that power users drive the Mac market, and the release schedule should be built around hardware and stability, rather than endless retreads and new consumer features we'll never use? I'm still on El Capitan and pretty much everything since then looks like bloat to me.
Whatever that is, it's definitely not a binary tree inversion bug.
You're right, I've just checked and when I set my cursor to be a custom color in accessibility settings suddenly my Firefox started to eat up memory as crazy :D. This went away as soon as I reset the cursor to default colors.
The author is specifically noting they're on Big Sur so they should not be affected by this.
Oh, you're right. I missed that.
That’s a nice catch, must have been satisfying to find this.
Also, it seems like a relatively easy memory leak to fix, hopefully in an upcoming OS update.
Looking forward to the fix.
Haven’t been bitten by memory leak issues yet, but I feel for you. I had an M1 mini for nine months (lightly used) and had no problems. Switched from an i9 MBP 16” to an M1 Pro 14” a week or two ago. Have had zero problems and it’s the best computer I’ve ever had and fastest new computer I have ever used. MATLAB is 2 times fast than my Windows desktop at work, which was substantially faster than my i9 MBP. Oh and MATLAB isn’t even native yet, but running under Rosetta 2. I also have never heard the fans, even with two 4K displays attached, which ran all the time on my i9 MBP. These chips and machines are incredible.
> I also have never heard the fans
We all do now
I'm also experiencing this on 1 of 2 M1 MacBook Pro Max 32GB machines. I've performed a full wipe/restore on the faulty MacBook and the issue persists. WindowsServer is usually the faulty process consuming 20+ GB of RAM. The only difference between machines is one uses icloud and the other doesn't.
Given the similarities between both machines and application usage, I can't help but suspect some type of hardware issue is plaguing a percentage of machines out there.
You didn't tell us the most important part -- does the machine exhibiting the memory leak problem use iCloud or not?
Assuming the iCloud-enabled machine is the one with the problem, it seems odd to jump to the conclusion it's hardware? That's a lot of extra code running on the machine; not just iCloud itself but every single iCloud-aware app. There's a lot of additional potential surface area with iCloud enabled.
The iCloud enabled MacBook is the one leaking memory. I'm leaning towards a hardware issue because the iCloud setup is the OS default and the number of reported issues is still low.
Edit: A recent reddit thread may have correctly identified the WindowServer leak being related to the amount of full screen video played. My impacted MacBook is often used for Netflix/Hulu content. https://www.reddit.com/r/apple/comments/qos5n5/memory_leaks_...
"I don't understand how virtual memory systems work, but I know that if I write an article complaining about <new Apple thing> I'll get a ton of traffic".
The Activity Monitor screenshot in the embedded tweet even explains what's going on. The system has 64G of physical memory. A bit less than 50GB is in use (so 14G is completely idle, wasted). 7G are wired (kernel, device buffers, etc.) and can't be paged. 38G are in used caching things for applications, and about 180G worth of compressed data (mostly zeros I'd say, since it's only using 5G of real memory) has avoided being paged out.
There is literally no memory pressure in this system, and no evidence of leakage in those numbers. Darwin deliberately tries to keep as much DRAM as possible live to minimise page-in and decompression costs; it literally can't find anything worth paging in or decompressing to fill those last 14G. That's a vacuum, not pressure.
I'm not going to debate the stability issues that the article author is complaining about; I do hope they have all of the crash reporting features enabled so that their crash data can be used by the various developers to help fix things.
But I would hope that before writing yet another stupid clickbait article, they pick a topic they know something about.
> 38G are in used caching things for applications
The 38G isn't "cached app memory" (there is no such thing), it's actual resident memory, calculated as the sum of the second column. The same as "RES" in Linux. There is no memory cache that differentiates apps from non-apps (whatever that is?), there is just regular file memory cache, which is labelled below the application memory section, shown to be using 13G.
Every file you open on your system (read: app you open, file that macOS accesses, etc.) will be automatically read into the file cache and left there until some event, such as not having any unused memory since all of it is used as cache, requires some contents of the cache to be evicted to make room for an application.
If this system had been low on RAM, this "Control Center" app would have been paging to disk to make room for the gigantic 26G it "needs" (the fact that it needs this much is a bug/leak. It should not be showing a number that high in that field, which is the resident memory required by the app.)
In summary, unused memory is meant to be used for caching files (see https://www.linuxatemyram.com/). The cache is lowest priority and almost considered "free" memory, in the sense that the cache can be evicted anytime to make room for actual application resident memory. In this case, "Control Center" is using 26G of resident memory, not cache (which is not shown in that column at all.) This is a bug/memory leak.
No, first of all, there is a screenshot in the article showing high memory pressure. That needs to be taken into account by your analysis of the author's intentions and understanding.
Second, I have one of the new new systems with a lot of RAM and when WindowServer takes up 20-30GB of RAM, and other apps take up 2-10GB more than they should, I see high memory pressure. There are other reports of this. Don't extrapolate too much from one screenshot showing a system shortly after reboot.
It's possible that most of the bloat from this bug gets pushed to swap and harmlessly stays there until reboot, to leave space in DRAM for actually useful application memory, but the typical user doesn't have guarantees of this based on the instrumentation available to them.
The second screenshot literally shows safari occupying 30+ G of memory. What the hell is it caching? The entire internet? Or maybe is it just memory leak?
I have the 32 M1 max and safari was using half the memory just opening YouTube. Must be something in the OS if it’s across lots of different apps.
Why would someone use safari when chrome is available
Battery life is much better with safari. Measurably.
They want to prevent Google selling all their data?
Google is no saint, but they don't sell your data.
Yeah, they retain it and give it to the government for secret mass surveillance subpoenas instead /s
Apple is no saint in that regard, but at least the data footprint is less.
for all of apples issues, i really respect their continued commitment to a personal assistant that doesn’t collect all of your data. as opposed to Am & Alph.
This is technically true, but only in a pointlessly narrow sense. They do not sell your data; they hoard it, and sell ad targeting services that use it. They also provide it to the government, for free, as required by law.
But does anybody actually care about this distinction?
It's tiresome seeing the same falsehood repeated in these forums.
If people want a catchy snippet perhaps: Google collects too much of your data, or Google profits from your data.
it’s tiresome to you because you are a liar fighting a tide of truth. if google pays your bills i can’t say that you should definitely be ashamed of yourself, but otherwise you should definitely be ashamed of yourself for trying to mislead people and be as duplicitous as possible at the top of your lungs. i find cultists really tiring.
Why would someone use chrome when Firefox is available?
Safari has been more efficient and is better integrated with the OS from my experience.
It’s significantly better for battery life, it’s a tiny bit more integrated into the OS, and as a bonus it doesn’t send every aspect of my entire browsing history back to Google.
And that tiny bit goes a long way if you use iCloud Keychain.
In before dang telling you to not create flame wars ;)
Why would anyone use Chrome when alternates are available?
Battery life, better responsiveness, less google spying. If you're not an extension power user and don't mind it's tab overflow behavior, it's pretty good.
That's an interesting point. I notice that the person who wrote the blog post is running Safari, too. Is this a Safari problem? The activity monitor screenshot makes it looks like Safari is eating up big chunks of memory for each tab. I don't often use Safari, and my M1 MBP has been problem-free.
A gmail tab consumes a lot of memory for me in safari but no int chrome. Maybe it has something to do with google products ?
Google meet reloads every couple minutes on my safari..
I have no proof but a strong gut feeling the root cause of this is broken OpenGL drivers. Here's why:
Over the past several months we've been very busy porting our Android- and Windows-based 3D Capturing app to iOS. We have quite a bit of OpenGL (ES) code that needed to be migrated. After a couple of initial hiccups things appeared to work well, however something was off. OpenGL code that's fast and completely unproblematic on Win/Android caused weird CPU stalls and after some minutes of app usage typically everything just crashed. After some digging and debugging it became clear that certain OpenGL function calls were leaking memory like crazy, as in several MB per call. Concretely I'm talking about glBlitFramebuffer() and glCopyTex(Sub)Image2D() . Both of these operate on frame buffers (memory that's often as large as your app window size) and they were internally allocating new buffers on every call and apparently never freed them.
Now of course this is iOS and the article is talking about MacOS but it's quite likely that Apple derived their M1-targeted OpenGL implementation for MacOS from the iOS version which has always targeted Apple SoCs.
Given the amount of OpenGL-based software still out there and Apple's deprecation of OpenGL in favour of Metal (they just don't seem to care any more) those kinds of large scale memory leaks don't really surprise me. OpenGL is still pretty much everywhere and if you have bad memory leaks in core framebuffer operations things start to accumulate very quickly.
Great perspective, thanks.
FWIW, I notice the memory leaks particularly with “Control Center”, which is an Apple extension to the task bar. I wouldn’t expect any Apple-written code to be using any OpenGL; it should all be in Metal.
Heh just killed Control Center with 41GB of memory...
Virtual or resident memory?
Do you know how long the Carbon API survived inside the Finder?
macOS still uses Carbon for much of the menu bar and internal event routing.
The Carbon framework is no longer supported or present on the mac. It went away with other 32-bit support in Catalina.
Yes, I see that now. Thanks. There seems to be a lot of confusion about what parts of Carbon are and are not going away or did not make the 64 bit transition. I haven't tried to build a Carbon app in a while. When you look in the frameworks, the ones that are present now seem to have arm64 bridge support, which I assume means arm native code for running Intel apps in Rosetta. Perhaps some very large user base app vendors like Microsoft and Adobe have not quite let go. Probably some old Apple code also. Apple keeps announcing it is disappearing but it can't seem to be killed.
No offense, but I laid out exactly what the framework is used for in my original comment: https://news.ycombinator.com/item?id=29144758. It's not (exclusively) there for Rosetta or third-party compatibility.
No offense taken. My mistake. None of my 32 bit Carbon apps have worked for a few years and I wrongly assumed the rest of it was gone.
If anyone wants to do any fun OS X GPU rabbitholing, this command will dump out a bunch of stuff from the IO Registry as an XML/plist:
ioreg -c IOGraphicsAccelerator2 -r -l -a
The PerformanceStatistics and IOReportChannels dicts are particularly juicy, and under IORegistryEntryChildren you can see which processes (IOUserClientCreator) are using what OS-level GPU resources, including Metal vs OpenGL API.
It's unclear to me if these issues are happening when you are running the iOS app on your Mac (emulating during development)?
If so, try setting Mac to use the discrete GPU all the time and see if it keeps happening. MacOS flip/flopping between discrete and integrated GPU was an issue for me. I can only imagine what MacOS might be doing during development of a graphics app such as yours.
I've been dealing with this ever since a software update to the M1 MBP last summer and it's been an absolute nightmare. It's obviously some kind of system-level memory leak, because my workflow is just basic apps (Superhuman, Chrome, Slack, Spotify, and WhatsApp are the only apps running most of the time; with the occasional other stuff like Keynote).
I used to do software development, running a million things at once and in decades I can't remember last time I saw the out of memory error. Sure, slow downs due to swap, but never straight up the error forcing me to kill apps.
On the M1 MBP, this error modal would pop up dozens of times a day. I would literally be constantly be quitting any app i'm not directly working on. It'd take a few days to get to this point, but at this point nothing other than a reboot would get it back to normal. The memory total for the apps running (plus the background stuff like Dropbox) was way below the built in 16GB, not to mention zero swap to SSD. Somehow there must be a leak in the OS.
Currently sitting at: WhatsApp Helper (Renderer) 3.57GB WhatsApp 1.37GB Dropbox 1.13 GB WindowServer 1.05GB etc.
This is not normal…
My main pain point aside from random OS crashes is Bluetooth / audio.
The sound device will just drop working randomly whether 3.5mm Jack or Bluetooth. Restarting the service sometimes works sometimes not, at first I thought it was MS Teams but didn't seem app specific.
Makes video calls painful knowing either sound or mic could just stop working at any time.
Why does this sound so similar to a Hackintosh? Honestly, I assumed people get a real Mac so they don't have to fiddle around with reboots etc
Bluetooth issues are often associated with endpoint security software such as jamf, crowdstrike and carbon black. If these issues are happening on your work laptop especially, you might have those things installed on them without you realizing it.
Apple's attitude towards endpoint security software is basically, "it's stupid and you shouldn't have them installed, our recommendation is to uninstall them" so that's that.
https://developer.apple.com/documentation/xcode-release-note...
Interesting, I do have jamf but no option to disable it due to corporate policy
Note that this also was when Apple released the EndpointSecurity framework specifically made to sandbox these sorts of corporate malware in user space and outside the kernel.
I've had my 14" M1 Pro for just over a week now. It has performed amazingly. But I also don't install tons of random apps and let them run in the background. I never restore from backups, I manually transfer files from my user folder over to my new setup. I always check whether an app has native M1 support before downloading, though I understand this isn't an option for everyones workflow yet. In cases where the app is literally a wrapper for a website like discord, I just use the website. Or in the case of zoom where I just don't trust the app to be on my computer and the website is sufficient. If something I used previous doesn't have native support I look for an alternative and a couple times I have found something better and/or open-source. I understand rosetta 2 has some issues with overusing swap, so I would rather avoid degrading my ssd and experience.
So far this has worked for me, and I'm able to get two full day's of battery life. I'm away from my desk far more frequently now.
Edit: My control center is using around 1GB of memory. Low enough that it wasn't causing issues but still concerning. It appears there is a memory leak somewhere, hoping a bug-fix is pushed out soon.
My M1 Mini crashes about once a week. I look through the crash log before sending it off to Apple, but is always seems to be something different that takes down the system. In comparison, my 2018 Macbook Pro rarely crashes and is running the same version of the OS.
I have the M1 Pro Max on my to-buy list, but I'm going to wait a bit to make the purchase since I really can't have my $4000 workstation crashing on me.
I'm kind of flipped around. My 16 inch 2019 MBP panics every two or three days. My 13 inch M1 MBP is rock solid and memory pressure is fine. Both running Big Sur. I won't call the M1 machine my favorite Mac of all time (my old 2013 takes that prize) but it's solidly in 2nd place right now. It's given me so little trouble...
My M1 Mini crashes (while sleeping) about once a week here as well. Funny thing is that I never get the crash reporter showing up after a hard reset. Very strange.
Two or three times, my M1 MBP has crashed overnight, but with the same error - something involving external drives. I keep 6-7 external drives connected at all times via two chained hubs. Hard to know if it's the OS or a flakey drive.
Beside that, it's been fine though.
16/512gb on my m1 mini is a great computer, but it’s not as ‘snappy’ as i thought it would be from all of the reviews. i run a lot of adobe.
Lightroom Classic (latest version) is unpleasant on my wife's M1 and even crashes. Last night I moved it over to my gaming PC which is pretty old, but performs much better.
It ran on her 2012 Intel MBP and whatever dayjob MBP I had in the past.
May I ask why you purchased the M1 model?
To replace the 2012 which was mostly OK, but had blown speakers.
I still have an Intel MBP provided by my employer. I don't keep a personal laptop, it would just gather dust 99% of the time.
First, thank you for your reply. I am curious, after this experience, would your next machine be also an Apple with M1 (or the next iteration of it), assuming nothing else goes wrong?
Probably, this is my wife's computer.
We aren't attached to Apple, but finding build quality, battery life, high quality screen and likely to last 10 years is hard elsewhere.
Thanks for the insight, I hope it works out for you all.
i’m hoping adobe has yet to fully optimize for the new architecture, and i’m also hoping they care to do so
It could also be that x86 has better SIMD support by a considerable margin, which can make repetitive memory access/serialized workloads a lot faster. I'm not super well intimated with those extensions, but I know that the NEON SIMD implementation in ARMv8 leaves quite a bit to be desired. It's a tricky situation, and one I don't see resolving in a nice clean way. It's stuff like this that makes me hopeful for RISC-V though, where we could theoretically have our cake and eat it too, with dynamic instruction pipelines and incredibly low power usage. Only time will tell, I suppose.
Throughput is not an issue on the M1, with 4x 128-bit SIMD units.
Neon is certainly not a bad SIMD ISA, it's a quite orthogonal one.
You also have the AMX extension at hand, which is more special purpose but allow to deliver very high throughput. (on a regular M1: 350Gflops DGEMM, 1.2Tflops SGEMM, without leveraging anything other than the CPU)
NEON can work pretty well and has been used in real time tasks on mobile for a long time. Example operations:
https://github.com/thenifty/neon-guide/
LR CC runs much much better on my M1 MBA than my prior 2017 MBP. I wonder if Adobe is just dragging their feet on updating classic since they want everyone moved to CC?
I wish it was just memory leaks on my M1 Mini!
Sometimes it just crashes. This has rarely happened on Linux. This sometimes happen on MacOS to the point where I don't trust it. At all. Ever. I have to save everything I'm working on constantly, just in case. Sometimes apps just stop working. Right now telegram just won't start. No matter what. Never happens on Linux. Sometimes it won't wake up from sleep. Never on Linux. Sometimes it loses the keyboard and/or mouse. This never happens on Linux. Sometimes the screensaver works, sometimes it doesn't. Grep is slow, waaay slower than on Linux. So are other random CLI things. How? Why? Docker... don't get me started, and don't make excuses. (~700 people at Mirantis and ~150k at Apple can't figure this out?) Everything runs slower on MacOS. Every. Thing. grep, sed, awk. Why don't sed, awk, grep, etc.. all the Linux things work quite right? I need to add something else on macOS to get them to work like they did on Linux. It's always "almost" but never do they work like I'd expect. I need to remember to turn off my monitor at night, because some nights it just randomly turns on. Never on Linux. Maybe it'll wake up, maybe not. I figured out if I unplug my headphones and/or webcam I can force it. All of these super basic things just DON'T work. It sucks. I could forgive Linux, but not an OS and Hardware made by the biggest company on the damn planet. I get it, this is faster than a 4 year old laptop. But that's not what I'm comparing it to. I'm comparing it to a 5 year old system I built myself from random parts that runs an open system operating system. That's a pretty low bar, and it doesn't win.
You might want to have your computer looked at. This is very much the opposite of my experience with the M1s at our offices.
So why use OSX? I've been running Linux for over a decade full time. I remember using osx at work and feeling like it was just clunkier, slower, and buggier.
What's the appeal?
It's not Linux. It's not clunkier, slower or buggier. Not sure why you had those issues.
I've gone back and forth between linux and macOS the past 7 years at work. Both get the job done, but there is absolutely no denying (in my experience) that macOS is slower.
Clunkier and buggies depend entirely on which distro you are comparing it with, though I find clunkier at least to be true.
Buggier than linux WMs and desktops is a pretty high bar to beat, as long as you're not including homebrew in the equation (worst, most necessary bit of software for Mac ever).
The parent poster is literally saying that OSX is all of those things relative to their Linux system, so I'm asking why they even have the OSX system.
> What's the appeal?
It runs the software I need. full stop
OSX is great, I love my M1 laptop. I think Linux is better but if you're on the go and not -too- into Linux that it is a fine substitute. All the Linux only coding that I need to do I can do in a terminal with remote login. At home I use Linux but on the go I have a work laptop with OSX and it's a good substitute, especially if you don't have to buy the laptop yourself.
macOS. It's not been called OS X for many years.
I get paid to use it.
Because last time I tried, I spent days to have fractional scaling on 4k monitors work without screen tearing, blurring etc. And have consistent momentum scrolling on touchbar.
I do not even think about these things on macOS or Windows.
At this point my time is too valuable for that shit. I used to fiddle a lot with my computers when I was 16. Now I want them to work.
i love linux but there’s no way macos is more buggy, I think my video drivers alone have more issues than apple’s entire os
Weird, I got an M1 Air recently and it's super fast (the most notably faster seeming new computer I've gotten in a decade) and all these issues are incredibly foreign to me. I just can't relate to this comment at all...
What use case do you have for your Mac? The benchmarks and battery life have been tempting me, but I use Docker a lot which I'm hearing mixed messages for, and to be honest, my Intel-based work Mac is less palatable to use than my personal Ubuntu setup, especially in regards to workspaces, and some weird Bluetooth issues.
I’m on the same boat. Using an intel Mac provided by work. Had an M1 MacBook Pro until a few months ago, and it was great at several things(fast, responsive, quiet, and great battery life) at the cost of crippled Docker use(I needed local docker at that time).
I feel like many many engineers like me would easily get a reliable laptop with MacBook Pro like specs. The problem is that simply does not exist.
Excellent hidpi display.
Ryzen 5000 series CPU with its low power draw.
Good build quality.
Reliable support system.
Non weird components such that a recent release of popular Linux distribution would “just work” without needing manual tweaks.
In my experience, you get 3/5 in the above list at best. Framework laptop hits several of those points but misses out on Ryzen. Another contender on my list is ASUS 14 OLED with Ryzen - I’m worried about their support system though.
If you can compromise on the HiDPI Screen, checkout the HP EliteBook 8X5 Series, they are amazing machines. You can plug up to 64gb of RAM into them if you need it, they run pretty smooth and my colleagues can run Linux on them without any issues.
Wow! Yes, this is exactly what I was looking for. Looks very appealing. I might get one! They're expensive, though. But it might honestly be worth it to me.
I want just to add that I am using a 8x0 g5 elitebook currently (because at the time there was not the 8x5 series which uses amd ryzens) and it's by far the best linux laptop I ever used in my whole life. Also very sturdy and looks great.
I'm even more tempted. My XPS 13 Developer Edition was also a very nice experience on Linux.
How is the trackpad?
Well you know, it's like the standard of the high end for windows laptop, still behind what apple put in their macbooks, but comparable with what you are gonna find on xps and thinkpad x/t.
Looks nice indeed. I have been spoiled by the excellent Retina display on Macs. For something literally in front of my eyes for a big part of the day, it must be good. I cannot see myself compromising to a lower res display :-( May be for others that are not as nitpicky as me….
Maybe way to got is simple linux laptop for work and an iPad for everything else. Would probably still end up being cheaper than a MBP.
Asus Zenbooks have always been solid on Linux for me, using them since they came out. Typing this on a UM425.
I run a Linux VM for development and it's glorious.
[link redacted]
how is the performance of the emulated vm? I would like to do something similar but I am reading mixed opinions
I don't have any numbers to give you, I can just tell you that the experience I have is that this setup performs noticeably better then my 2019 MBP.
I had some issues with Docker when I first got my M1, but then I realized I was using an amd64 image. Images built for arm work fine for me.
If you can get it built for arm. I've got so much build errors I didn't bother to check.
Docker is hopelessly broken on the Intel Macs, and from what I’ve heard it’s certainly not any worse on the M1 Macs.
> it’s certainly not any worse on the M1 Macs.
Probably not any better either. If you want to run docker on a mac you probably should not, but the second best strategy is probably to bite the bullet and run an actual VM in which you do your docker stuff.
Odds are a proper and well maintained hypervisor you keep running will be a lot better than whatever DfM does.
> Probably not any better either.
Probably, although you probably do still get better thermals and battery life with the M1 Macs.
> not any worse
About 6x worse perf in my specific case.
Fair enough. My Intel Mac gets about a 5x performance penalty doing Webpack builds in Docker versus natively. My coworkers using M1 Macs report similar numbers. What workflows are you seeing M1 Macs getting 6x worse Docker performance than Intel Macs?
sort of explained here https://news.ycombinator.com/item?id=29144096
Eh, it is worse. Worse to the point that even though I just got a MaxBook, I won’t even conceive of getting rid of my 2020 maxed out Intel iMac. And it’s why I have other x86 machines too for Linux or whatever, because containers on macOS are rough but containers and Docker specifically on the M1 is really rough. It is clearly an ecosystem/use case that Apple doesn’t prioritize.
It makes the case for stuff like GiHub Codespaces or sshing into a remote machine via VS Code a lot more manageable. For everything else, the M1 is a beast, and even emulated, you can get a really good Linux experience (Windows too for that matter — games, obviously not withstanding), but containers? Nah. It’s worse than the already sad state on Intel Macs.
Yeah I'd carefully look at what your expectations are with regard to compatibility. Docker itself seems to run just fine but running X86 (AMD64) containers is pretty slow compared to normal macOS docker which is already slow.
I've loved my M1 MBA but I purchased it specifically knowing I have a primary x86 workstation to fallback on for heavier tasks and I'll often give up trying to install some dep/run some program on the MBA and switch to that. It still works great as a general purpose device and I do some dev on it but if you're looking to do all your dev on it, I'd be cautious.
Def got slightly burned by docker on M1 Pro. Inherited a python project with 250 deps. Building docker image (part of is compiling spacy) used to be ~10 minutes, now it's a whopping 65 minutes.
Personal programming projects and normal web browsing type stuff. What has been most noticeably faster for me is compiling code.
Apple refuses to ship gpl3 software for seemingly no sane reason (judging by the fact other commercial software vendors including e.g. windows don't have this issue) probably explains the command line issues; apples' tools are an entirely different codebase from the tools in linux distros. And unless you actively compare systems, you might never notice the apple versions are different.
It's been this way since OSX day 1 IIRC, so it's a little surprising the apple versions are still apparently inferior (if the grandparent poster is anything to go by, haven't used em in years), but it's conceivable the CLI just isn't an apple priority.
I wish they'd just give up already and copy as much (of the CLI-relevant stuff) as possible from linux, and accept that they'll need to give up a tiny amount of control over a small sliver of their OS - a sliver that they never wrote in the first place and also can't really change the api of, because that would break all kinds of things, which renders their ability to close source somewhat moot anyhow. And most users never use this stuff! The linux-conventional gnu variants are well maintained and usable in both MS world and linux world - having CLI scripts that are more easily portable across three OS's would be great. Apple's toxic control-freak nature really hurts here, and mostly just their users. I really don't get it.
Anyhow, there are workarounds (but e.g. homebrew is hit and miss), it's mostly just an annoyance they're even necessary.
It’s a perfectly reasonable reason: they were forced to open source their objective-C compiler that was built on gcc and decided they didn’t ever want to be in that position again. The BSD tools are annoying, if you’re used to the Linux versions, but they’re perfectly reasonable implementations of the relevant POSIX utilities (better than busybox, for example).
What? NeXT open-sourced their Objective-C gcc extensions in 1992, a half-decade before Apple bought them.
Yeah, it’s a Steve Jobs thing. As far as I can tell, this is a significant factor in the rise of LLVM.
https://news.ycombinator.com/item?id=9158017
It didn't help that rms was using it as a "hah hah that time I put one over on Steve Jobs" a decade later when I saw him speak. No idea if he still does.
I sure hope apple has a better reason for shipping unnecessarily impractical software than some pet peeve by a now-deceased CEO. That would be kind of insane.
> jordigh on March 6, 2015
> They're basically achieving their dream now with Swift, though. I don't expect them to ever release that source. I don't think there is any precedent for Apple releasing source code after the fact.
I wonder, to what degree is this still the case today? Has Apple's position on Open Source changed in the past half+ decade?
I don’t think it’s changed about GPL: GPL3 made things even less friendly to what they want, so they’ve started doing things like switching to zsh instead of bash. LLVM/Swift are open-sourced, but not GPL
And here we should note that LLVM (by its creator who already worked on it at Apple at this point) was offered to GNU and the FSF in 2005 but were ignored/chided [1] until a decade later when GCC/GNU/FSF realized what they had been offered ten years earlier and then had to grapple with what they had rejected, once LLVM had taken hold [2, 3], including proposals to try to refuse LLVM patches to emacs, out of some sense of petty NIH thing [the whole thread 2 and 3 are part of].
[1]: https://gcc.gnu.org/legacy-ml/gcc/2005-11/msg00888.html
[2]: https://lists.gnu.org/archive/html/emacs-devel/2015-02/msg00...
[3]: https://lists.gnu.org/archive/html/emacs-devel/2015-02/msg00...
Reading those threads makes those decisions seem pretty reasonable to me; this doesn't look like a petty vendetta at all. And while chris sounded optimisitic about including LLVM, he didn't offer to; he offered the glue code to allow the gcc frontend to talk to the separate LLVM backend - and doing so implied a raising a number of fundamental requirements, such as the C vs. C++ bootstrapping issue.
Accepting major changes like that isn't trivial; and a certain amount of NIH is healthy here. I mean, that's kind of apples motto too right? This was a technical patch set with unclear upside (if any), and potentially relevant cost - why would they accept that?
No, GPL2 made them do that, and even so they were ok with it and used to pay people to work on GCC for the Mac in the old days. They rejected GPL3 specifically.
> they were forced to open source their objective-C compiler that was built on gcc
A better phrasing: They were unable to close-source an existing GPL'd compiler.
I think the above phrasing is fair, and technically correct. You can closed-source a fork of GPL software. It's extremely common for companies do this while software is in development. It's the moment that someone distributes the software that the license requires source distribution. Some people/companies never distribute their forks. This is why newer revisions like AGPL were written.
Probably because of the 'provide the keys' and not sue for any anti-circumvention tech in the GPL v3. The anti-tivo provision. Not sure if it applies to standalone system utilities, but I bet it is an internal blanket ban on the GPL v3.
Apple is very fond of locking down their hardware. Google took a very strong stance against the AGPL-even banning it from their own public code repo service before they shut it down.
If you like to add random crap from home brew all of the time, you’ll notice different issues than you did on x86
My M1 Air crashes everytime I wake it from sleep and then immediately watch a Youtube video in Firefox. I can reliably reproduce this. If I open the laptop so it wakes up, login, and open a Youtube video it will hard crash close to 100% of the time.
Firefox just changed how their full screen video playing works in the most recent released (version 94). Could be that. Try an earlier version of Firefox.
M1 Air here that I've had for almost a year with zero issues, and I also use Firefox and watch a lot of youtube.
This is literally the opposite of what most people claim. Does anyone else have these issues? I actually just ordered an M1 MacBook myself.
M1 is great, but you can certainly hit some issues where only an x86 version is available and the M1 version is either not there or buggy as hell.
I’ve had a couple crashes with by 2020 MBP, but my experience is that it’s the best Mac I’ve owned since about 2010.
I can't speak to m1 in particular, but BSD grep is known to be very slow regardless of platform. You can just brew install grep to get a fast one, but you'll still always have to always deal with avoiding the slow system grep.
ag and rg are other fast options.
The speed of GNU grep vs BSD grep is discussed on the BSD mailing lists here: https://lists.freebsd.org/pipermail/freebsd-current/2010-Aug...
Ripgrep changed how I search my codebases, it’s so damned fast I love it. I almost always use it when I’m trying to find a pattern or a bit of code I vaguely remember but have forgotten where it might be in a large codebase. Still returns almost instantly.
Some mac users seem to be cursed with bad hardware.
It seems to be black and white: users either have no problems, or ALL the problems.
I started to get "the curse" after 8 years with my 2012 MBP. It started randomly turning off. Not crashing, just turning off! I replaced the battery because that was the only warning I was getting. But it can no longer run any macOS. Even mavericks. It crashes randomly. It went from rock solid to fragile as an elderly lady with osteoporosis rollerskating on a cobblestone path.
Now it is my primary Ubuntu machine since the touchpad works so much batter than Ubuntu running on my Lenovo Yoga and it has better performance. Zero turning off problems. It's been on for months running a VNC server.
Yeah, it is very strange. I have an early 2014? MacBook Pro and it still runs today. 0 issues. That being said, it's now slow as hell and likely because of the battery. But the development experience on MacBooks is delightful.
Don't know about others but my M1 Macs are the most stable and performant Apple machines I had in a long time. My previous 16" i9 MBP was a disappointment (laggy/sluggish, running hot) compared to my M1 Mini and MacBook Air.
I also don't see any memory issues, etc.
What most people say is not always the truth. Eg; "Windows 10 is the last Windows."
Most people don't give any sources.
I have all the issues the parent post is commenting about. I thought I was crazy... but same happens to me. Simple stuff... why doesn't sleep to wake work correctly? Why are my peripherals schizophrenic? My Windows 98 Compaq didn't have these issues... and that was two decades ago.
> Why don't sed, awk, grep, etc.. all the Linux things work quite right? I need to add something else on macOS to get them to work like they did on Linux.
If you mean how they work, like command line arguments and such (and not just speed), that's because macOS is based on BSD and thus using the BSD and not GNU versions. That's not really a "not right" thing, it's just different, and would be the same if you used any other BSD like OpenBSD, FreeBSD, etc.
Yep... GNU versions are available for Mac, i.e. `brew install gnu-sed` and then `gsed` works like you'd expect.
If you install GNU coreutils from Nix (and maybe some other package managers, idk) you can get them with their normal PATHs, at the cost of some possible incompatibility with macOS-specific scripts that assume the bundled versions.
I used GNU ones without the `g` prefix for a year or two and never ran into trouble because of it.
no, it's definitely "not right" -- mac os ships with ancient, crippled versions of the bsd utilities, afaict simply because it's bad for open source and because they can. E.g. `man sed` then check the date at the bottom
Mine says March 27, 2017 and it does everything I've ever asked it to? Maybe you have different standards for what qualifies as ancient and crippled.
For me (on Catalina), it's May 10, 2005.
Yes they rarely update the utilities for non-security reasons, but they do update them. I'm on Big Sur and it's March 27, 2017 for sed.
How often do you think utilities like sed, awk, and grep are being updated? They're virtually unchanged for 20-30 years.
I'll bite and talk about centos, the 'stable' distro.
Take the example of 'tar' that slowly evolved and some time even changed its default behaviour. e.g. what to do when untaring a folder but there's already a symlink with the same name as the folder, do you erase the symlink and create a folder or do you untar the folder in the symlink? Between centos 5 and 7 this changed... Causing me some amount of pain.
hostname stopped supporting '_' in host names. Could have raised a warning, have added a '-pedantic' option but no.
hostname stopped assigning the permanent host name with the 'hostname myhostname' command and now you need to use hostnamectl. Could have left the 'old' feature and make hostnamectl whine instead, but no.
ifconfig the deprecated tool subtly changed the order of fields too and now the MAC address is not in the same place as before, breaking scripts that trusted that not to change.
df changed its syntax, adding or removing a new-line some time ago.
Yeah thinking back we shouldn't have used the standard tools and expected them not to change.
So, now, whenever I see in code review some bash person using 'standard' tools to get info that could be gathered through procfs or sysfs, I instead ask they use procfs, sysfs or ioctls and be done with it.
Hmmm, maybe "everything is text" wasn't such a good idea after all...
Powershell might be what you're looking for. Text is thrown away in favor of .NET objects (and reflection)
Yes, I'm a fan of PowerShell! (I just wish it was a Linux thing and not a Windows one...)
To be fair, the last two are more from a bad design choice by *nix in general to base everything on passing around unstructured text and hoping each tool's ad-hoc schema won't change.
That 'hope' was true for some time, I guess. Those systems had no problems with previous migrations from early fedoras to centos 5.9... When your last line of compatibility is procfs and sysfs and ioctls or posix calls (statvfs for free disk space), well I know what Linux devs think of their installed userbase's time. Yeah yeah free stuff for free alright. Still, if you'd paid for rhel, same shit. Thank your dogs we provide an /etc/redhat-release, heh.
And if course even Linux procfs is not without its surprises. Adding columns or lines was foreseen but some changes are more fun!
Take the evolution in /proc/meminfo that changed the meaning of some of the lines there over time (WTF?), whilst adding others (which was fine, expected, etc.). Of course I want now to include IO cache (that will be reclaimed if I ever needed to malloc) in my system's free memory indicator! Heh you can still get back the old value with this formula, what are whining about?
Gnu grep uses Boyers-Moore for searching - it's wicked fast. BSD grep installed by default on Macos might not be using this algorithm.
This 2 year old HN post shows the performance difference between Gnu grep and BSD grep on Macos:
https://news.ycombinator.com/item?id=19522456
> How often do you think utilities like sed, awk, and grep are being updated?
sed 4.8 - 15 Jan 2021
awk 5.1.1 - 28 Oct 2021
grep 3.7 - 14 Aug 2021
What am I missing here?
June 10th, 2020. What did I do wrong?
macOS ships ancient versions of GNU programs which are now GPLv3. The BSD things are kept up to date.
If only they had a few billion in spare change to write their own workalikes.
The Telegram issue is not specific to M1. I think it's a bug in Telegram, as it simply refuses to start on my Intel based MBP at the moment. Started this evening.
Funny, I've been feeling the same way about my M1 MBA. It's hard to deny that the newer silicon doesn't have my older Linux rig beat for latency, but when it comes to raw performance I was pretty damn disappointed. At risk of offending some people, it was exactly what I feared the M1 would be: a phone CPU shoehorned in to a clamshell body. There are advantages and disadvantages to running a machine like this, but overall I don't really find the development experience to be that much more compelling than a Raspberry Pi.
That being said, Apple still knows their audience. I imagine there will be a lot of chronic Twitter browsers and Pages advocates who fall in love with this machine, and I think that's fine. I do have to wonder what constitutes a "premium" device if not for the hardware though. Everything I do on modern MacOS feels like it's prodding me to go spend $8 on a tool that I'd have for free on Linux or even Windows. Booting up Big Sur makes me long for my old a1502 running Mojave and handling some frankly ridiculous workloads while staying rock-solid in terms of stability. If anything, I applaud Apple for coming straight out and telling me that the future of their devices isn't for me.
> That being said, Apple still knows their audience. I imagine there will be a lot of chronic Twitter browsers and Pages advocates who fall in love with this machine, and I think that's fine.
That is delightfully insulting. Real developers would never use Macs, of course.
Real Developers use whatever and still get shit done.
If your M1 experience isn't better than a Raspberry Pi (the only issue I have on RPI is speed and memory limits) then something was wrong with your laptop. There is really no comparison, the M1 will destroy it in any performance demonstration.
But when it comes to raw performance I was pretty damn disappointed.
I'm curious what you mean by "raw performance"? I'm no Apple enthusiast, but AFAIK the M1 is the fastest laptop CPU on the market. There are faster desktop CPUs, but I don't think that makes it "a phone CPU shoehorned into a clamshell body" — it beats pretty much any laptop CPU in existence in the benchmarks I've seen.
It's hard to find software native for Windows and Linux simultaneously, yet unavailable for macOS.
> Everything I do on modern MacOS feels like it's prodding me to go spend $8 on a tool that I'd have for free on Linux or even Windows.
Yeah, that’s so annoying. Wanna browse a .zip file (not extract it)? Either `brew install file-roller` with its non-macOS UI, or use shady paid shovelware. Wanna record your desktop audio as part of a screencast? It’s so complicated [0] and often involves shady stuff [1] or software that costs $99 [2]. What about Windows? There’s (mediocre) ZIP browsing support in Explorer since Me, 7-Zip is free and trustworthy, and OBS just has a desktop audio input built-in (and I also could do that easily in 2005). Linux? You’ll probably get a working archive tool built-in (file-roller, ark, whatever) and OBS desktop audio just works.
[0] https://billykorando.com/2020/04/14/capturing-desktop-and-zo... [1] https://www.youtube.com/watch?v=H4VPbMvBuLg [2] https://rogueamoeba.com/loopback/
> Wanna record your desktop audio as part of a screencast?
It doesn't have to be a screencast, it's kind of a convoluted process regardless of which operating system you're using [1]. If I'm not mistaken OBS is using the built-in virtual loopback devices available through pulseaudio, but MacOS does not include a virtual loopback device by default, hence requiring a third party one such as Blackhole or loopback.
> ...and often involves shady stuff
...what shady stuff are we talking about here? Blackhole [2] is open source software. It's about as shady as any of the (several) linux audio systems.
[1] https://en.wikipedia.org/wiki/Loopback
[2] https://github.com/ExistentialAudio/BlackHole
> Everything I do on modern MacOS feels like it's prodding me to go spend $8 on a tool that I'd have for free on Linux or even Windows.
Amusingly, this is exactly how I felt when using Windows.
Memory management on iPad seemed to be terrible when I used one for a while. An app would start acting up, so you would take a screenshot or video and send to Apple's bug reporting tool right? Every time you switched from the bug reporting tool to any other app (to crop the images of the bug, etc.) it would run out of memory and kill the bug reporting tool, which wouldn't save any of the text you had entered in the forms, deleting all of your work on the bug report.
Also, every browser tab with text entered was a random crap shoot if you were going to lose it by going to another tab to copy something and return to paste. Nothing was predictable, youor work was just constantly lost of the device. Almost no apps seemed to persist things when the OS forced them to shut down, except maybe the highest rated dedicated artist and note taking stuff. But even Apple's own apps (like the bug reporting tool) often didn't bother.
I'm strongly prefer Linux (been using it last 10 years), but I spent last year on macOS in waiting for Asahi to become usable. I still curious about your experience.
Did you actually tried to replace Apple garbage version of command line utils with latest one from GNU? It's kind a obvioius that OOTB version of utils will be much slower since Apple stopped updating GNU versions once license been changed to GPLv3. And BSD version of utils they ship with newer macOS is also usually inferior.
I don't think this is actually super obvious to someone that isn't familiar with macOS and is used to using linux. I do think that if someone finds something peculiar enough to notice and complain about it that they should at least do their due diligence to ascertain why, rather than simply being upset about it.
As much as I would like to get one of those M1 Macs, this is exactly the reason why I will wait for the 2nd generation of their processors. My Intel MacBook runs rock solid and stable. Until the ARM Macs haven't reached that level of stability, I am not willing to make the switch.
Rule #1 on new mac hardware: wait til the 2nd generation!
Rule #1 on any new hardware, wait for 2nd gen. E.g. computers, cars, aircraft.
And by 2nd gen, I mean first update after a major redesign.
M1 Air here that runs rock solid and stable. OP is talking nonsense. They should get it replaced under warranty.
This might be the case for some users and basic use cases like surfing the web and text writing. But there are still a lot of issues for pro users and developers. If you rely on such a machine for your daily work, its not mature enough.
> Why don't sed, awk, grep, etc.. all the Linux things work quite right?
Why should all Unix implementations work just like Linux? What's so special about Linux? Why doesn't Linux work like BSD?
Also with various package installer you can sub in all your favorite linux versions of things. I have l.env file that aliases them all to linux equivalent commands.
Because the "linux" versions tend to have more features and better help output? I've never seen anyone install Linux and be like "damn, I wish my coreutils were more obtuse" and install a bunch of non-GNU stuff. Like, I get that Linux is just another unix and the GNU userland isn't somehow more legit than others, but it's not like people are intentionally getting inconvenienced just to be annoying.
Or Linux users are self-righteously declaring their way is the only way...
This is usually called "embrace, extend, extinguish".
> tend to have more features and better help output
Just a random opinion. Others disagree and prefer the BSD way. Concrete reasons include code-base simplicity.
https://github.com/coreutils/coreutils/blob/master/src/yes.c
https://github.com/openbsd/src/blob/master/usr.bin/yes/yes.c
That's a developer-perspective issue. You're talking to a user here, and from a user perspective, the GNU utilities are generally strictly superior.
Take the example you provided, of yes(1). Most of the extra code relates to optimizations, and ensuring that the entire argv is echoed back rather than just the first argument. Because of this, even a very simple example like:
"yes hello world"
will behave differently on BSD vs GNU, with the GNU version likely being what users expect.
> with the GNU version likely being what users expect
Lol why? Not everyone grew up on GNU.
> strictly superior
This isn't something you can quantify and say 'strictly superior' it's all just taste and opinions.
Because the functionality in this case is a strict superset, regardless of what you're accustomed to.
I didn't grow up on GNU either, but the GNU yes does everything I expect having used other unix systems.
> This isn't something you can quantify and say 'strictly superior' it's all just taste and opinions.
For a complicated tool, sure. But this tool is very simple. The GNU yes(1) accepts 100% of inputs that work on the BSD tool. It does a few additional things too.
> Because the functionality in this case is a strict superset
Extra functionality is always better?
Then I can fork GNU, add extra junk options, and give you a strictly superior tool? Nonsense.
The GNU tools are the Homer Cars of UNIX. Not everyone wants that.
Hey, look. As a systems guy myself I understand the aesthetic around clean and simple software. That's a totally reasonable preference.
Here's the thing: GNU userland isn't some poorly designed aberration. It's by far the most popular unix userland in the world and is very commonly ported to systems which may do things differently. Almost every other unix has a distro to supply GNU userland tools on it -- no other unix userland can make the same claim.
Not everyone wants it, sure. Especially people with perspectives broader than the typical user. Of course. But it's the most popular system in the world -- not some weird never-used prototype.
And in the case of yes(1), I think you would have to try very hard to find a reason for a user of the tool to have a negative preference.
> Then I can fork GNU, add extra junk options
You likely can't, all the possible option flags already have some meaning. <:)
edit: I can't read
> OpenBSD `yes` doesn't look very useful if your language isn't English, or you're dealing with any other kind of prompt than y/n
Not sure what you mean - you can get it to print whatever you want by passing that on the command line?
Technically Linux is “just another unix” but macOS is an actual certified UNIX
There are certified builds of CentOS/RHEL
People don't realise this because they don't use the name 'Linux' anywhere.
https://www.opengroup.org/openbrand/register/
If you really want to get technical, Linux is just a kernel. There are certainly implementations of it that are POSIX-compliant, but since most people use it with GNU systems it's a bit of a moot point.
> Docker... don't get me started, and don't make excuses. (~700 people at Mirantis and ~150k at Apple can't figure this out?)
The whole reason Docker exists makes complaining about its performance on non-linux operating systems asinine.
> Why don't sed, awk, grep, etc.. all the Linux things work quite right?
Because MacOS is BSD-based and those tools aren't "Linux" things. You can install the GNU equivalents easily if you dislike them.
> I could forgive Linux, but not an OS and Hardware made by the biggest company on the damn planet.
Software development doesn't particularly care about the fact that Apple is the biggest company on the planet. The pooled resources going into development of linux is quite substantial and yet it too still has plenty of software issues. LTS kernels can have hundreds of patches because software development is hard no matter how many eyes you have looking at it.
> The whole reason Docker exists makes complaining about its performance on non-linux operating systems asinine.
TIL there's a single "whole" reason Docker exists.
Please, enlighten me, what exactly is that reason?
A quick search would answer your question: https://www.docker.com/why-docker
Can you distill that into the single whole reason you're referring to? It's unclear to me just reading that marketing page.
If it's "to provide a standardized unit of software that allows developers to isolate their app from its environment, solving the 'it works on my machine' headache", I don't see how it makes expecting good performance asinine.
Linux's ABI is stable, MacOS can definitely implement it... there's nothing in Docker/Linux containers in general, forcing MacOS to incur virtualization overhead.
It doesn't solve the "it works on my machine" that's the ironic thing, that I've discovered through someone else too
Docker is built on specific features of the Linux kernel. From Wikipedia https://en.wikipedia.org/wiki/Docker_(software):
> When running on Linux, Docker uses the resource isolation features of the Linux kernel (such as cgroups and kernel namespaces) and a union-capable file system (such as OverlayFS)to allow containers to run within a single Linux instance, avoiding the overhead of starting and maintaining virtual machines.[12] Docker on macOS uses a Linux virtual machine to run the containers.
And MacOS is built on Darwin; a hodgepodge of FreeBSD atop Mach and years of iterating on that.
Point being, Apple has zero barriers to making Mach look like an existing interface to userspace a la FreeBSD. They can take advantage of Linux's stable ABI and add the necessary Linux interfaces for native Linux containers support, and it could be performant.
The whole x86 on ARM problem seems like the more difficult one to make performant from where I'm sitting, but there's no shortage of claims to the contrary saying x86 emulation on the M1 is fast enough even for demanding video games.
>Please, enlighten me, what exactly is that reason?
To serve as an easier to use facade on top of Linux kernel virtualization options. It only works on Linux - or with an emulated Linux.
It truly is a shame we have to choose between quality hardware and a quality OS.
My maxed out 2013 Macbook Pro (the last model that runs linux without issue) has finally worn itself out and I’m in the market for a new laptop, but I can’t find a new laptop worth spending money on. I’m currently doing all my work on my desktop just hoping I don’t have to travel any time soon.
Check out the Dell Mobile Workstations. I had to buy a laptop and went with a custom 7750
Look at Framework. Really fantastic out of the box Linux support (some WiFi issues depending on kernel version but legit it’s better than what you’ll find on identically specced Dell's right now and you have options about brining your own preferred WiFi card), excellent Windows support, a good community, repairable — I’ve had mine since September and am a huge fan.
I think you convinced me to not get a new MBP. These are the exact kind of issues I had with prior MBPs. Also Emacs was always inexplicably 5x slower than on Linux. It's a shame we can't use alternative OSs on the hardware.
I'd swear I'm hearing myself talking about Windows 20 years back... "I keep saving everything incase it just crashes", "It's slow as hell even for simple tasks", etc. Has Apple 2021 become Microsoft 2001 ?
Not really, but you should see Microsoft these days...
Microsoft is actually quite decent these days, my desktop hasn't crashed in like 2 years
My kids were giving me a hard time because they cant work out why I hit Control-S every 2 minutes.
What you describe is not a Mac Mini with some OS or HW bug, it's a broken one. Don't know what your issue is, if it is hardware, or it was pwned, or has 2-3 incompatible Rosetta apps wrecking havoc. But I'd get it replaced.
None of these things happen to people, much less altogether.
>Everything runs slower on MacOS. Every. Thing.
The experience of everybody who bought an M1 is everything (or almost) running far snappier.
>Everything runs slower on MacOS. Every. Thing. grep, sed, awk.
So, is it "everything" or is it "grep, sed, awk"? And what is your case exactly? Benchmarked them on the mac and on some Linux system with the same file? Maybe did it on a network attached drive?
>Why don't sed, awk, grep, etc.. all the Linux things work quite right? I need to add something else on macOS to get them to work like they did on Linux.
You seem not to understand this POSIX system or versioning. Apple has older (or BSD) versions of those tools. Use Brew or similar and install the GNU versions, which would be like you know them from Linux. Not that the BSD versions are broken, they're just different.
Let this comment be a reminder to everyone that "you're holding it wrong" is not a response solely used by desktop Linux advocates.
Yeah, it's also used by any other kind of person, if your experience is not representative to the point of being bizarro.
Expecting the GNU and BSD apps to be identical is certainly understandable, but explaining that they are different is hardly the same as “you’re holding it wrong.” It’s information. If, instead of learning from it, you want to be mad at that information, let alone the messenger, I pity you.
I was more referring to the intro paragraphs:
> What you describe is not a Mac Mini with some OS or HW bug, it's a broken one. Don't know what your issue is, if it is hardware, or it was pwned, or has 2-3 incompatible Rosetta apps wrecking havoc. But I'd get it replaced.
> None of these things happen to people, much less altogether.
> The experience of everybody who bought an M1 is everything (or almost) running far snappier.
Especially since there are reports of the same thing as described happening qll over the comments here.
I did not give due credit to the rest of the comment, it was quite well written. I apologize for that coldtea.
Putting aside your dislike for BSD versions of userspace tools, I am confused as to why it hasn’t occurred to you that you may have defective hardware or a defective OS install? Have you undertaken even the most basic troubleshooting steps?
A lot of those weird random things would happen to me on Linux. But I agree with you, I wouldn’t expect a $2K+ vertically integrated, 10th iteration flagship laptop to have these kind of issues. It definitely does suck. I have the 16” base model M1 Pro and I’m also considering a return
> Why don't sed, awk, grep, etc.. all the Linux things work quite right? I need to add something else on macOS to get them to work like they did on Linux.
This is the GNU in GNU/Linux :)
I really like GNU coreutils, and other basic GNU utilities like GNU sed, find, and awk. I know some BSD folks see them as ‘bloated’ and ‘non-standard’, but to me they're plenty fast and they feel complete.
When I used macOS for a while for a job, it surprised me how much I'd come to rely on the features and behavior of certain GNU utilities that are apparently peculiar to them. But I miss 'em when they're gone.
I already have these issues on my 2019 MacBook Pro. Presumably it's the OS rather than the hardware but who knows
I still don't understand why the majority of people haven't switched to Ubuntu or Kubuntu... They're quite user-friendly nowadays and they perform amazingly well. I often go for months without restarting my Kubuntu laptop. It performs exactly the same as it did years ago when I first bought it. What I like most about Linux though is that it's open source. It actually feels like my laptop.
I have Windows installed on it as a dual-boot and I cannot believe it when I log into Windows (Windows is definitely worse than Apple)... I just don't understand WTF happened over the past decade; Windows seems to have gone backwards. There is no way in the world I could ever use this thing and be productive. The interface is so clunky with title bars which use up all the valuable screen real estate. Difficult to switch between different windows. The cursor was almost impossible to use with the touchpad (it just feels unresponsive); astonishingly, I was actually able to get the touchpad working well with Kubuntu with the generic open source touchpad driver which came with it (it's extremely customizable).
Somehow, the narratives around open source software being inferior quality always seem to be disproved in the long run. While the quality of enterprise software tends to degrade overtime (as big corporation financial politics and bureaucracy creep in). Open source projects are free from toxic financial incentives so they tend to keep improving over time.
I love my Regolith/Ubuntu setup and don't want to move anywhere.
However, I can't find a decent machine which will last for 4 5 hours on battery, and it is important. My brand new Thinkpad T14 lasts less than 3 hours, and I am on minimalistic setup with only few apps running. Tried all kinds of power management tweaks.
OTHER thing is the issue with the microphone on external devices in Linux which is pain, which is why I always have to use my phone for calls.
"I write for a magazine and web site called MacWorld, yet I'm unable, unwilling, or both, to really look in to the problem. Let me complain and hope someone from Apple reaches out."
Why is it his job to debug the issue? Raising awareness so that Apple prioritizes it is useful too.
This is caused by macOS 12 Monterey. Having the same problem, but seeing reports from Intel Mac’s as well.
From the article:
> That’s extremely similar to the MacBook Pro and Monterey issues I’ve read this week—except I’m still running Big Sur.
I missed that.
Maybe there is more than one issue here, but Intel users having the same issue on Monterey where they did not have it before seems telling. Maybe M1 machines had a different OS build or features that caused this issue in Big Sur and now it's universal in Monterey?
This may sound random, just my gut feeling for debugging kicking in. Since Control Center seems to be implicated in this, and Control Center opens ports for Airplay connectivity[1], has anyone tried turning off Airplay sharing? Or just turning off wifi/connectivity completely and seeing if the memory pressure lets up?
[1] https://developer.apple.com/forums/thread/682332
I have Airplay disabled on both of my machines and only one has been impacted by the leak.
My M1 Air gets a ton of abuse and I don't see any abnormal perf issues. There was a super annoying bug where the laptop would crash when I closed the lid with a bluetooth mouse connected, but that's it. I'm still on Big Sur, maybe Monterey is buggy.
Apple has bunny hopped hardware architectures, successfully. The time is right to switch from the 70s FreeBSD software environment to Bell Labs's 9P concept-oriented operating system jazzed up with Apple's delightfully intuitive UX Pixar styled art and animation. Linux is a step sideways from the BSDs.
That sounds like a whole lotta work for very little gain.
The speculating about the OS arch made me wonder:
If such an issue were to occur on iOS without a crash, would anyone even notice? There's no real way to monitor RAM usage, and the failure mode is less clear (slowdown, issue with responsiveness, followed by likely reboot).
iOS would likely kill the app as soon as it got out of hand, and rely on the state restoration hook to bring back the state as the user left it. Normal macOS apps don't have that, though.
If the leak was in some part of the system, maybe a problem, but I'm sure they test pretty rigorously, and get enough crash reports to figure things out fast with that many devices in the wild, all in pretty constant use.
It may have to do with many of the new Mac (Apple-authored) apps being Catalyst apps. Not sure. I feel that Catalyst is a bit of a "stopgap" solution, so it may not be as thoroughly optimized as a more permanent solution. I'll bet something new comes out, once the entire line is switched to MX chips.
For myself, I just got the new M1Max 14-inch MBP, and it's quite nice.
I had to remove my Thunderbolt docks (I have a range of docks, including ones purported to be "Thunderbolt 4") from the video chain. I have found that the computer seems to lose track of what is where, if I have a dock in the loop.
Catalyst is the stopgap before SwiftUI. Eventually we'll see all Apple apps migrated to SwiftUI from AppKit, UIKit, and Catalyst.
SwiftUI is still better called beta software at this point, to the point where 1password abandoned it mid rewrite to go to electron. It's gonna be a few more years at least.
Considering that many corporations are just, now, beginning to take up Swift, after it has been out for seven years, I'd say this was a pretty sure bet.
I like SwiftUI. It shows a lot of promise, and I look forward to releasing products, written in it. I don't think it's especially mature, and maturity is important, when you need to release ambitious projects.
I'm not really in a hurry, though. I'll be getting around to working in SwiftUI, and I tend to do well at the things to which I apply myself.
Yeah, possibly, but seems 50-50 that that Catalyst stopgap is until Apple thinks iPadOS plus some virtualisation is good enough to replace macOS.
The browser taking up so much memory are probably just object/memory leaking advertising scripts. Run an adblocker and watch the memory usage of your browser tabs go down significantly!
Nope. As pointed out, Firefox had a nasty bug that was interacting with an Apple bug that causes memory leaks. Firefox has fixed the bug but from their comments it seems that it was more of a workaround around the Apple bug rather than a fix.
> I now keep Activity Monitor open while I work
I also do this, but I wonder is it a bad idea? Is Activity Monitor enabling any extra logging that itself is slowing things down further, or is it just subscribing to already enabled events?
I'm on M1 MBA 16GB and I get the out-of-memory errors too. It's still a world of difference from the 2016 MBP I upgraded from.
As touched on in another comment, the problem might be the 8 desktops and 20 Firefox windows I keep open.
I've always tried to avoid keeping Activity Monitor open. On my current 2015 MBA, it doesn't eat up a lot of memory, but it usually does list itself as a top consumer of CPU cycles, often 15-20%. Running `top` in the terminal only eats about 2%.
It would be ironic if Activity Monitor itself were causing a memory leak.
I'm a long-time Windows user and it's customary to keep Task Manager open (more for its tray icon) to keep an eye on what the system is doing. Is that not common in the Mac world?
We have apps like istat menus that behave similarly. Activity Monitor doesn't create a menubar icon that lets you monitor system state like the windows equivalent, but it's dock/taskbar icon can be set to behave the same.
I also think windows task manager is fairly efficient and TBH more useful with their full system state graph view. You have to click one resource at a time to see system state of your memory, cpu, etc and there is no GPU graph like windows.
Opening it from time to time, sure, keeping it open, no. Power users may install a menu bar applet to track the stuff but you generally don't need to bother.
Honestly unless something seems to be sweating the box there's not a great reason to monitor it all the time. By far the biggest CPU/memory culprits are web pages, but just looking at a bunch of chrome or ff processes doesn't really help determine which it is. Activity Monitor's own GUI is a pretty serious offender, running around trying to allocate icons and graphs for all those apps. If you just need to know what's going on, `top` on the command line will watch it for you. But usually (from a 2015 mac perspective) there's no need to keep an eye on it.
I did notice that CalendarAgent uses around 350MB which is ridiculous to just sync my Google Calendar. When you kill it, it restarts itself goes down to 15MB but gradually goes up. Doesn't seem to go above 400MBish, but still... Other than that, no complaints. Other apps seem to manage memory rather well.
Rather than working with Activity Monitor open 24/7, just install iStat Menus. Makes it really easy to monitor your system.
I have a MacBook Pro (13-inch, 2018, 2.3 GHz Quad-Core Intel Core i5, 16 GB 2133 MHz LPDDR3) and now on macOS Monterey everything just feels sluggish (especially the animations) and crashes sometimes.
It just feels the OS is designed for different, more recent hardware. I have been using Mac since 2010 and that wasn't the case in the past. I paid about $3,000 for this device. I am not happy and considering Linux/Windows.
Reading this with mild concern as I'm waiting for my first M1 Max, a couple questions spring to mind...
* I wonder if this is more prevalent when the hard drive is approaching full capacity? It would be interesting to see the breakdown between physical and swap when it happens.
* Has anyone observed these seemingly random memory leaks while not running anything under Rosetta 2?
There is apparently a bug in the buffer cache, that can be alleviated by regular purges:
https://macperformanceguide.com/blog/2021/20211029_0832-digl...
I'm facing the similar issues with my m1 air(8 GB). Recently firefox managed to get to 44GB. I'm closing and re-opening my applications; and checking the memory pressure when things start to feel sluggish. Also unrelated but my speakers have started to sound like garbage.
a) You need to raise an issue then with Firefox team as that isn't something that Apple is at fault for.
b) If your speakers sound bad then it could be dust or debris. Might want to unscrew your Mac and use a blower / vaccum cleaner on it.
Not just a MacBook problem but Apple M1. So far nothing “bad” has happened when my Mac runs out of Ram.
Interesting hypothesis that it’s because MacOS doesn’t handle the memory structure well. I would have expected this to have manifested under either arch or neither.
These situations are interesting. My M1 13” will occasionally get a stuttering cursor that only goes away with a restart. I haven’t had this problem in over a decade on any computer.
I know I’m not unique in this because of the number of threads on the Internet on the issue.
Interesting to see the odd interactions between software and hardware. Though this sort of bug does put me in mind of how in Minecraft when Notch fixed light transmission through glass, it would start raining through it too.
I had this in Gnome about a year ago, and it was a Gnome memory leak by an extension causing it.
Funny. The last time I had this was in 2014 or something when I had too little RAM and the system started swapping. Linux GUIs perform poorly under memory pressure. Way worse than windows or MacOS for some reason.
Been a constant experience for me from 2001 when I first started using.
It's not just GUIs. Running out of memory on a Linux system is really bad even if it's a headless server otherwise under light load
I think even quantifying memory pressure on Linux is relatively new
https://www.phoronix.com/scan.php?page=news_item&px=Linux-Do...
macOS' behavior under high memory pressure is much more graceful IME
My Intel-based iMac (32GB) has been struggling in Safari lately (more than usual) with stutters and general poor performance; there was a Safari update not too long ago, right?
THIS. Safari has been running like junk since the last update.
Been using an M1 mac and Big Sur on the daily for a year and have never experienced this, and I have the 8GB model where I would notice it. My Control Center is using 29 megabytes.
I think this specific memory leak is only in Monterey. I definitely have to manually kill Control Center every few days otherwise it'll be using 20GB+ of RAM. Supposedly it's fixed in the 12.1 beta though, so hopefully Apple releases a fix for it soon.
The author specifically mentions that he is using macOS Big Sur: "That's extremely similar to the MacBook Pro and Monterey issues I’ve read this week — except I’m still running Big Sur."
It really highlights the value of the fleet-wide profiling that ChromeOS enjoys. There’s a dashboard somewhere of what’s using all the space (and time) for ChromeOS users in the wild. But HN just actually hates telemetry more than it hates memory leaks.
I'm running 12.0.1 and Control Center is using ~65 MB, with an uptime of 12 days. This is on an M1 MBA. What type of machine do you have? I wonder if there's some specific software that causes/exacerbates the issue.
Seconded. M1 Air, 8GB, Big Sur and no such issues for ~1 year.
Maybe a leak in a Max/Pro-specific kernel component? I assume Control Center depends on some hardware-specific kernel modules.
What I find common on such threads is Linux users bashing how Apple users are suckers. I don't get the point of such comments. Most people on HN can already figure out how to build their own PC, install and configure the distro of their choice. But thats not even relevant as everyone is pointing out the flaws in software so that the company responsible will fix them. Bragging about Linux makes no sense, many of us using Macs here most likely already have a linux box as well.
Zed Shaw posted on Twitter his findings about Safari memory leak here in this thread [1].
Seems to be related to video tag and according the comments here people are on YT or other sites that have videos embedded.
[1] https://twitter.com/lzsthw/status/1457387347700588550
interesting, I just discovered a Touch Bar daemon process consuming > 4 GB of memory on my 8GB MacBook, a clear leak.
I've been dealing with a weird memory leak on my Windows desktop for a out a year now. For some reason, dwm.exe (this seems to be the window manager) uses more and more dedicated GPU memory. This is really weird since it uses so much that apparently Windows does paging of GPU memory.
I wonder if Apple's hardware people are going nuts from the fact that they created a world record chip and its potential is being wastes by subpar OS.
Further, I wonder how much of their vast resources might Apple be pouring into making a new OS without these problems.
I started with an 8GB M1 Air and it was unusable.
But I also regularly leave dozens upon dozens of tabs open and use multiple heavy IDEs...
Upgrading to 16GB was enough to fix it instantly and I no longer have the freezing/crashing issues, so I'm surprised 64GB machines are running into it
I recently got an m1 mini, just on principles the 16GB version, and did already get the RAM usage warning once, with my browser going into large memory usage. So there may be something to this story, will watch for it.
Huh I have an 8GB air and I run Kubernetes locally and develop a fairly complex system with Skaffold on it (~10 services). Thing is a beast, had no problems at all, but one does need to be aware one is not on the 128GB-of-RAM workstation :P
I mean how much demand will a locally hosted dev service generally have? Nothing about K8s inherently denotes high ram usage.
Browsers are generally the biggest culprits for memory usage. Random tabs will leak memory and before you know it some news article page is using 1GB, your GMail is using 1GB, etc.
Actually my co-founder who is a tab-hoarder has had a few issues. I am a tab minimalist so you're probably right!
It does sound like you expected more performance out of the 8gb model than was reasonable.
There were plenty of people claiming things like "8 GB on M1 is like 16 GB on x86/Windows" ...
I feel like this claim was made largely by nontechnical YouTube influencers and some Mac fanatics talking about how fast the swap on the SSDs is and equating that to not needing more RAM.
I don’t think it’s a wholly unfair assessment. You can get away with more when the memory management stack is smart and fast, but in the end if you need 9gb of ram and only have 8gb you’re going to know it. Most people just don’t (at one time).
I didn't have much of a choice, at the time the 8GB model was the only one you could get without a lead time.
So I got it knowing as soon as the 16GB dropped in stores I'd replace it
No problem. Ram is very cheap nowadays. Just get a stick of 16 gb DDR4 ram and slide it in.
Having this on my 13 inch Macbook Pro M1. After upgrading to Monterey it also started to heat-up without any heavy app being open. In Big Sur I had to open say OpenTTD to make it warm. Now it's just browser and IDE doing that.
Is it specific to M1? On my Intel Macbook Pro I have to restart Safari daily to fix memory leaks that make the address bar incredibly unresponsive (even closing tabs doesn't work - I have to quit and reopen the app).
My hot reload JS dev server crashes on my M1 16GB between 1 to 7 times a week. This has never happened on my Windows 10 desktop machine (16GB too) with the same code base. My M1 is still ~15% faster though.
I dunno, my M1 with 8GB is not even really usong swap afer a month of uptime and heavy loads memory-wise (VM's mostly).
Based on other comments: try not using chrome. I don't and I have 0 issues iwht this.
Interestingly I have seen this on mm _Intel_ Macbook Pro - but _not_ on my M1 Macbook Air.
I've noticed worst applications for this are those built on the Electron framework (MS Teams, Slack, Evernote).
Managing shared memory is not a new thing for Apple, as it has been doing the same on iOS for a long time. Maybe a bug in Monterey? None of the M1 users I know had this issue.
Hm. I got my M1 Max on the morning of 30 October...
:) so no issues here.
Been using M1 Pro for a week now. Docker (amd64) runs waaay slower. App crash (mostly Safari) at least once a day. Haven't noticed memory leaks.
> There are also lengthy Hacker News and Reddit threads with users recounting their experiences.
For many years, the press kept HN a secret and wouldn't even link or mention it as source. Presumably to keep HN from falling into mainstream. There were lots of times you could tell they got it from HN, or even screenshot without mentioning the site.
But for the past 2 years it seems more and more mainstream media are giving direct links.
I could only wish the first rule for mainstream media is to not talk about HN. But then I guess it is too late.
Happens to me all the time on my Macbook Pro too. I hate this operating system. What I would give for a Apple hardware + Linux machine.
There wasn’t any mention of how much free hard drive space this machine has, which I believe is more relevant than ever now, isn’t it?
Might be totally off base and unrelated, but I’m curious how often people are power cycling and if that provides any benefit
I once had Telegram using 120gb of ram (swap).
Telegram has occasionally done this on my Intel Mac mini.
I think the way it caches media is bugged right out. I changed the cache size preference from 32gb to 8gb and it eased some of the pressure on swap, but it was still insanely high. I had to switch from the App Store version to the "Desktop" version you download from the website.
Perhaps a coincidence, but Telegram in a Firefox tab also uses a phenomenal amount of ever increasing swapspace. I've seen Firefox reach 67GB with a Telegram tab open, it is distinctly noticable compared to just about everything else I open in the browser.
Still an issue in latest 12.1 beta. Really frustrating. Just dropped a chunk of change on an M1 Pro too.
Lol the timing of this complaint couldn’t be more timely with the arrival of new 14-inch MBP…
Jumping in to point out that while this stuff happens on linux, at least one doesn't feel so hand-wrigingly powerless when it does. I know it's a choice between starting a career in maintenance and complete learned helplessness.. and the former option well seem a lot more appealing when it's no longer available.
Ironically this horrible website is itself a memory pig -- holy crap ... the ads ...
What ads?
Once more, hardware is little use without good programming...
I use Firefox on M1 Air with only 16GB of memory and have had zero memory issues. I have like 100 tabs open and no problems. Machine and OS works perfectly and is rock solid. No issues at all.
Definitely noticed sporadic freezes on my M1 Air.
No signs of the issue on my 16GB RAM M1 Air. On Monterey, installed on day one, zero reboots since.
I haven’t had any issues. I sometimes run two VMs at a time on my 13” MBP with 16GB RAM and although the memory usage is almost maxed most of the time it’s been running fine.
I'm not seeing this on my M1 Mac Mini
For the past 10 months, I’ve used a 13-inch M1 MacBook Pro supplied by my employer.
How is this possible ? The M1 MBP is only available since October 26.
Not true. There was a 13inch M1 MBP available. It just was not the redesign that has non usb-c ports, etc.
M1 MacBook Pro (13 inch) and M1 MacBook Air were released near the end of last year. The newest machines (14 and 16 inch) and the M1 Pro and M1 Max MacBook Pro. The combination of "MacBook Pro" and "M1 Pro CPU" can be confusing. The original poster appears to have gotten an original M1 CPU in the 13 MacBook Pro about 10 months ago from their employer.
It’s Dropbox…
you are the beta tester
Looks like all those Big-O-focused sdev interviews really paid off!
O(64 x 2^30) = O(1), so it's still constant memory usage.
And that is why I won't be buying one until at least 2025, and unless they also remove all on-device spyware and let me power it off for real without it broadcasting its own location when shut down.
Apple's great hardware is crippled by the piece of shit operating system called macOS.
Man up Apple and release Apple Silicon drivers for Linux.
I know you are venting, but seriously, I use macs for macOS. Otherwise I’d be buying much cheaper hardware.
Mac hardware is nice, but it’s not that nice. Except for maybe the trackpad.
Apple Silicon has a lower power draw than any comparable AMD or Intel chip.
I'd like to use one of those Mac Minis as a homeserver/homelab and macOS is the only thing standing in the way. Similarly, I doubt there is a non Apple laptop that matches the M1 Macbook Air's battery life to performance ratio, specially in that form factor.
I know Asahi linux exists but reverse engineered drivers will never be as effective/performant as official drivers. Porting drivers to Linux won't even be expensive for Apple. They refuse to do it just cause theyre Apple.
I don't think there is anything else where you can run that much GPU power on battery as well as wired in a form factor anywhere close and with integrated system/gpu memory. The AMD APUs haven't gotten close yet, but as things start moving to LPDDR5 maybe they will.
I personally haven’t been able to find another laptop with the same battery life, design, trackpad, and screen quality of mac hardware.
I bought an M1 max laptop and I’m almost tempted to go back to my Linux machine because of several issues I have with macOS. There are things that are by design like mandatory animation delays for things like switching spaces and going in/out of fullscreen, which are borderline dealbreakers for me given how often I do those things. And there are also things I suspect are either bugs or straight up performance lapses, such as scrolling through text in a terminal at 4k being slow if you have ligatures enabled.
I could go on, but I think it’s a perfectly valid sentiment (regardless of whether or not they were venting) to wish for mac hardware with a Linux OS.
You might find it helpful to turn on System Preferences > Accessibility > Display > [x] Reduce motion. I go through phases of preferring the fades it applies instead of certain animations, although it still doesn't get it to "instant". For terminal, iTerm offers better choices for you to determine the balance between perf and display settings, incl a GPU renderer.
The Reduce Motion setting isn't anywhere close to good enough.
Please don't post generic-indignant comments, or take HN threads into flamewar, or call names in comments here. That all pushes discussion in the direction we're trying to avoid here.
https://news.ycombinator.com/newsguidelines.html
Memory leaks are a feature of a software, not hardware. I read the article and while the Apple and Adobe software can be blamed, there is nothing in this post that it would pin this on the new hardware and the author does not seem to be technical enough to understand this.
It might be in Rosetta 2, or in the M1 builds of some apps.
Still not hardware
Apple hardware and software are closely linked, it's not like you can use Rosetta2 on any other hardware of your choice.
Rosetta 2 is an essential feature of M1 Macs due to their hardware, that is not is not a feature of Intel Macs.
There’s nothing saying that the ARM macOS port isn’t to blame. Porting operating systems is hard.
Totally agree. I read the article as, when I am on my M1 I often experience memory pressure related issues. Not, because of my M1 I often experience memory pressure related issues.
It's very likely future software updates can help address these issues.
Honestly, the machines haven't had wide deployment in the field for all that long for Applications as huge as Photoshop. It takes time to discover, fix, and redeploy things like Photoshop. And it's not like Photoshop never leaked memory on an x86.
> There’s nothing saying that the ARM macOS port isn’t to blame.
There is. People have been running ARM macOS for about a year and I haven’t seen this complaint before.
It's obvious you haven't read the article and are reacting to the headline. In fact, the author highlights that the issue is probably with the system software (macOS):
That’s extremely similar to the MacBook Pro and Monterey issues I’ve read this week—except I’m still running Big Sur ... It’s possible that macOS isn’t managing this unified memory structure properly, and will continue to allocate RAM beyond what is available without freeing up RAM that is no longer needed. This is commonly referred to as a “memory leak.” Performance gradually deteriorates until you need to either wait for the RAM to clear, force-quit the app, or restart the machine ... it’s entirely possible that Apple fixes it in a future version of macOS without ever addressing the issue publicly. That’s what happened with the excessive SSD usage earlier this year, though that Apple says that was a “data reporting error” and not an actual problem. The memory bug is an actual issue that needs to be fixed as soon as possible.
It's a fair criticism that the headline is misleading, even if the text is accurate.
The big thing is that M1 chips have RAM in the same package, which allows for performance improvement if the software takes advantage of the new architecture. It turns out that the system they came up with offers good performance but doesn't tolerate memory leaks very well.
It is made worse by the fact M1 MacBooks are limited to 16 GB of RAM, non upgradable of course.
So yes, it is a software problem, and it will get better, but the M1 is particularly sensitive to it.
> It turns out that the system they came up with offers good performance but doesn't tolerate memory leaks very well.
Can you be more specific here? It seems natural that a system with limited memory doesn't tolerate memory leaks very well.
It is just a wild guess based on the article, but the unified architecture is mentioned. So maybe each subsystem (system RAM and GPU in particular) all have direct access and therefore more of a tendency to fight for resources and maybe also complex and therefore more error prone locking patterns over shared memory.
Why doesn’t it tolerate memory leaks very well? The fact that the system remains responsive and able to close the application indicates that it does tolerate it well. I’d say macOS in general tolerates memory leaks much better than Linux does, which if you don’t have a swap disk and you run out of memory the entire system just stops functioning.
Don't worry everyone, I'm not experiencing this issue. -- Apple flans
Please don't post unsubstantive comments here.
can some mod do something about this user "dang"? he seems to crapflood every discussion I try to read with these stupid judgemental comments and never contributes to the discussions, doesn't hacker news employ moderators of any sort? I know I'm a new user and don't have enough "karma" or whatever to downvote but this guy is one of the most irritating and useless commenters I've ever seen, and that comes from someone who has used reddit and slashdot for quite a bit.
I'm one of the moderators here (and currently the only one who posts publicly as a mod). Sorry that wasn't clear.
If it helps at all, the comments you're complaining about are even more tedious to write than they are to read.
https://hn.algolia.com/?dateRange=all&page=0&prefix=false&so...
not sure if that's some in-joke copy-pasta, but dang IS a mod, here.
One wonders if there is some weird software his employer installed? Probably not. Apple just dropped a new chip into a laptop designed for another chip and its not a good match. Mac controls the entire production process including software. That laptop was designed for a different chip. Everything else with the M1 is fine because the hardware was designed for the chip