We now have a cpuidle driver, which significantly lowers idle power consumption by enabling deep CPU sleep. You should also get better battery runtime both idle and during sleep, especially on M1 Pro/Max machines.
Thanks to the cpuidle driver, s2idle now works properly, which should fix timekeeping issues causing journald to crash.
Also thanks to the cpuidle driver, CPU boost states are now enabled for single- and low-threaded workloads, noticeably increasing single-core performance.
Thermal throttling is now enabled, which should keep thermals in check on fanless (Air) models. There was never a risk of overheating (as there are hard cutoffs), but the behavior should now more closely match how macOS works, and avoid things getting too toasty on your lap.
Random touchpad instability woes should now finally be gone, thanks to bugfixes in both the M1 (SPI) and M2 (MTP) touchpad drivers.
A bugfix to the audio subsystem that fixes stability issues with the headphone jack codec.
New firmware-based battery charge control, which offers fixed a 75%/80% threshold setting. To use this, you need to update your system firmware to at least version 13.0, which you can do by simply updating your macOS partition to at least that version or newer. This new charge control method also works in sleep mode.
U-Boot now supports the Type A USB ports (and non-TB ports on the iMac), so you can use a keyboard connected to any port to control your bootloader.
And last but not least, this kernel release includes base support for the M2 Pro/Max/Ultra SoCs! We are not enabling installs on these machines yet as we still have some loose ends to tie, but you can expect to see support for this year's new hardware soon."
"New firmware-based battery charge control, which offers fixed a 75%/80% threshold setting. To use this, you need to update your system firmware to at least version 13.0, which you can do by simply updating your macOS partition to at least that version or newer. This new charge control method also works in sleep mode."
This is interesting, am I correct in thinking this a feature implemented by Apple and now supported by the Asahi team? Does that mean that macOS supports this charge control feature?
I really hope Apple brings the same charge limiting to iPhone as well.
> This is interesting, am I correct in thinking this a feature implemented by Apple
Yes, battery charge control is a hardware(/firmware) feature supported on other modern laptops as well, such as the Lenovo ThinkPads, but it's not a standard so it requires explicit driver and OS support.
OpenBSD recently added support for this as well for both of these implementations (Apple silicon and ThinkPads).
In my experience for it to take effect you need either a predictable window of plugged in time (it will delay charging until it needs to to hit 100% by when you normally unplug) or just to use your compute almost exclusively plugged in (at which point it will start letting the battery drift down to 80%). If you regularly unplug to work on the couch or whatever you may not ever see the effect directly unless you wake up an hour early to check the charge state of your laptop.
I really want to try this out on my work ThinkPad since it stays plugged in effectively all the time. It requires a kernel module though, I'd need to get that approved first and I haven't got around to it.
Naïve me doesn't understand why it can't be done from userspace, but I'm sure there is a good reason...
If your laptop firmware doesn't support it, here is one trick if you have a removable battery:
Put a piece of paper over one of the batteries middle contacts. That will make the firmware think the battery is overheating. It will then refuse to charge, but will still happily discharge.
You can do that to keep your battery at 80% while still on AC power. Handy if you operate from AC power 99% of the time, yet don't want your battery to die from being stored at 100% charge and hot for many years.
> Put a piece of paper over one of the batteries middle contacts.
Bad, very bad idea if you don't know what you are doing - depending on where the "smarts" in the BMS are, you may damage your battery or make your BMS think the pack is broken or prevent your BMS from recognizing a charge state mismatch (and in the worst case, a cell going undervoltage or reverse polarity) as you have a good chance that you cut off one of the cell balancer contacts. This trick only works with removable phone batteries.
I'm a corporate firefighter assistant. If there's one thing people should not be messing around with, it's Li-Ion batteries. These things are dangerous enough even without people messing around with their control and management systems.
> I really hope Apple brings the same charge limiting to iPhone as well.
This was added to iPhones in 2019.
> If your iPhone stops charging at 80%, it's most likely due to a feature Apple introduced in iOS 13 called Optimized Battery Charging. It aims to prevent over-stressing the battery and hence extend the battery life of your iPhone by limiting the charge to 80%.
Your iPhone learns your usage patterns and delays 100% charging until moments before you wake up in the morning.
It's a bit different, though. I don't often need my iphone to last multiple days on end. Yet, if I keep it plugged in as often as I'm sitting at a desk, it'll never go below 80%. If I get it below 80%, sooner or later, it will figure "i want to use it" and will charge it all the way to 100%. The lowest my battery ever got on this phone was 40 something when I was away for a weekend without a charger. It's very rare I use it a lot, so the "intelligence" clearly doesn't care how long the battery needs to last.
The way it's implemented on my mom's android, it always shows 80 or 85% (can't recall which one it is), even if she leaves it plugged in for the whole weekend.
On my HP laptop, if I activate the "battery saver mode" (as opposed to "AI"), it reports the maximum capacity as somewhere around 80% of the design capacity. I don't know whether Linux cooperates with this, but probably not. HP only talks about OS compatibility for the "AI" mode, which not only requires Windows but a specific HP app.
Sure, but if I know I won't drain my battery more than 10% that day, I can't tell it not to top it up more than 80%. That's the case for me 99.9% of days.
This also seems to work only if you've drained the battery below 80%. If it says 90% and I plug it in? It'll charge it fully right away.
It doesn't charge the battery over 80% at all, unless it detects that you have a normal waking time, in which case it charges fully right before you normally wake up.
Unfortunately, other companies copied the 80% charge bit without copying the part about figuring out if you have a normal waking time and giving you a full charge right before that.
For instance, Samsung's S23:
>Once you turn off the battery protection function, you'll be able to charge your battery up to 100%
> It doesn't charge the battery over 80% at all, unless it detects that you have a normal waking time, in which case it charges fully right before you normally wake up.
My iPhone disagrees with you. I go to bed and wake up at just about the same time. If I somehow managed to drain the battery below 80%, when I plug it in, it'll charge to 80 and then tell me that by the time I wake up in the morning, it'll be fully charged. Which is the case.
But the most common use case, for me, is waking up, leaving the phone plugged in (WFH). Maybe go for a walk around noon, snap a picture or two. Forget to plug it back in and get back to work. Plug it in as I go to bed at 95%. It tops right up, doesn't wait until the morning.
> Unfortunately, other companies copied the 80% charge bit without copying the part about figuring out if you have a normal waking time and giving you a full charge right before that.
But that's actually what I want. With very, very rare exceptions, I never need a full charge on my phone. Hell, until a few weeks back, I was rocking an iphone 7 with a battery in a questionable state (started bulging). That thing never went below 60% with my use patterns. 60% meant night out, so a hefty dose of maps use. Normal days, it didn't go below 85-90%.
I think the issue is that Apple expects that when you wake up in the morning, you unplug your phone for the day and actually use it a significant amount. Which isn't my case at all. If I don't leave my house, it'll stay plugged in. I usually forget about my phone since everything I need to do, I do on my computer. For the occasional phone call, I can either keep it plugged in (the cord is long enough) or it's already connected to my headset, so don't even need to go fetch the phone. If I don't go to work, it can stay plugged in for days on end (I don't always grab it when I go out).
Because it's my understanding that charging and keeping it all the way full shortens significantly its battery life compared to 80-85%. I would rather not have to dick around plugging and unplugging the phone to not have it be full all the time. Plus, that also counts as consumed cycles.
I've used my old phone for a good six years. I've swapped its battery some three years ago, and it would've needed a new one now, had I continued using it.
If that phone is any indication, my current one should be in service for at least as long. If I can avoid having to swap its battery, it's a win in my book.
It can also sometimes happen that I foresee being away from an outlet or otherwise need as much charge as possible. In those situations, I'd just deactivate the battery saver feature and let it charge to 100%. So if the remaining capacity is closer to its original one, again, it's a win.
> If you don't charge fully until just before the user's normal wakeup time you aren't keeping it all the way full all the time.
You keep repeating this and seem to ignore my observation that if the battery isn't drained below 80% when I plug it in, it will recharge it fully immediately. It will not wait until the user's wake-up time.
And in my case, it's rare that the battery falls below 80%, so whenever I plug it in, it gets recharged fully right away.
So, in practice, it's all the way full all the time.
My Sony Xperia 10 IV lets you set it to never charge above 90 or 80 %, as an alternative to it learning your habits. I have it set to 80%, and I've been unable to use it up in under 2 days. I've heard that iPhones have similarly good battery life as the 10 IV, so it seems, to me, that get quite far on 80%.
I bought the same phone recently, and for my light usage pattern (some texting, the browser, Google Translate, Google Maps, a calculator app, and the camera), the battery lasts almost 5 days!
I had completely forgotten how it is to own a phone that doesn't need to charge everyday...
> This is interesting, am I correct in thinking this a feature implemented by Apple and now supported by the Asahi team? Does that mean that macOS supports this charge control feature?
It does, but in a weird way. You can turn on "adaptive charging" and it will randomly decide to charge to 80%.
If you want to properly control it, just install the wonderful AlDente utility ( https://apphousekitchen.com/ ). Then you can manually control the max charge percentage. Mine is permanently set to 80% because I never really use even 40% of the battery on my M2-based laptop.
>And last but not least, this kernel release includes base support for the M2 Pro/Max/Ultra SoCs! We are not enabling installs on these machines yet as we still have some loose ends to tie, but you can expect to see support for this year's new hardware soon."
> e includes base support for the M2 Pro/Max/Ultra SoC
Does this mean Apple gave them prerelease hardware early? Might apple start helping these guys more - like for example donating a 5 person dev team for a few months maybe?
It means that Apple isn't radically changing the internals of the SOC every year.
>Apple’s first iPhones ran on Samsung SoCs, and even as Apple famously announced that they were switching to their own designs, the underlying reality is that there was a slower transition away from Samsung over multiple chip generations. “Apple Silicon” chips, like any other SoC, contain IP cores licensed from many other companies; for example, the USB controller in the M1 is by Synopsys, and the same exact hardware is also in chips by Rockchip, TI, and NXP. Even as Apple switched their manufacturing from Samsung to TSMC, some Samsung-isms stayed in their chips… and the UART design remains to this day.
M2 Pro/Max were available in January. I think they needed to wait until now to be sure the M2 Ultra announcement didn't have too huge of changes from the way the M1 Ultra was done. In other words, the Asahi Linux team don't have an M2 Ultra to test on, they are getting ready for when they can get some test results, possibly from users.
Asahi Linux's site previously redirected any site with the Referer header set to 'news.ycombinator.com' to google.com. They did this because they believe that Hacker News' community fosters a hostile environment, and thus it was blocked. Setting `rel=noreferrer` on an `a` tag means that the browser will not submit this header, thus meaning this sort of redirect will not occur. Also(as per Marcan's Mastodon account), this rel tag has only been added to asahilinux.org links, not any other links on HN. This means that the moderators of HN believe that the commentary that has in the past been quite misinformative should be allowed, and they are deciding to bypass the block that was put in place by the Asahi team. Now, as per https://social.treehouse.systems/@marcan/110503331622393719 , they're adding an HTML header to anyone who has submitted a post to HN(or has the submit page in their history), as they have no other way of doing anything, as they no longer have the Referer header.
This raises interesting moral questions that I'm not sure I have an answer to.
For one, I don't believe this place fosters a hostile environment, although it's definitely a place where people love to tell you how you're technically wrong about something.
For another, I would guess that there are a very limited number of websites that would opt into this sort of anti-traffic behavior. Hacker news could certainly choose to honor it, but it also feels within their right to bypass the block.
I wonder if there's a sort of middle ground, where HN alerts the user of the redirect that would have occurred, but still shoots the user to the desired location.
But now you're getting into user flows and begging the question as to why the redirect is there in the first place.
I'd love to know more about the perceived hostility, even reading up on Mastodon left me with more questions than answers.
there is constant transphobia against Asahi contributors on display on this site. Every single article that gets to the frontpage will have dozens of vile posts in the deads and a couple sprinkled inbetween. I’m not sure if HN staff doesn’t care about it, agrees with them or simply isn’t aware.
It’s not limited to the Asahi project, either. The site isn’t safe for queer folks.
I had a look at the comments on this post. I don't see any saying anything at all about gender, pronouns, trans-ness, etc. (Other than in this little discussion right here about what is allegedly wrong with HN.) I have showdead on, incidentally; I'm not seeing any dead comments on those topics either.
Maybe for whatever reason this particular post is better than others? I put <<asahi site:news.ycombinator.com>> into DuckDuckGo. (Not because there's anything magic about DuckDuckGo search results, but to get a sample of Asahi-related HN posts without cherrypicking.) The first hit is to https://news.ycombinator.com/item?id=35394297; no trans-related things there that I can see either. (It does have some discussion of why the Asahi Linux site was turning away people coming from HN, and the reason given there was nothing to do with "Lina", nor with transphobia, but was that marcan feels that HN discussions of Asahi are full of mistakes.)
(To look for trans-related things I (1) skim-read the comments by eye and then (2) searched for "trans", "gender", "pronoun" and "man", the last because people being obnoxious about trans issues often can't resist going out of their way to call someone a "man" or a "woman" if they think they might be upset by being called that. I specifically kept an eye out for dead comments. It's very possible that I missed things, but it seems to me that a minimum threshold for saying that there's "constant transphobia on display" is that someone explicitly and somewhat carefully looking for it should be able to find at least one example.)
Next few links aren't obviously strongly-Asahi-related articles. Next one that is is to a comment about "Lina" somewhere in the middle of https://news.ycombinator.com/item?id=35233479, so I took a look at the comments there. There was indeed a "Lina" subthread, some of which was pretty rude but none of it in a visibly-trans-related way.
My general impression of the HN comments on Asahi-related threads, from this, is that they are on the whole very positive, that some people think marcan is weird for Lina-related reasons (which may or may not actually make sense; I have no knowledge of that business), and that if anyone is being obnoxious at or about trans people in those threads then either they're doing it in ways I'm failing to see or else it's being cleaned up effectively enough that it's gone by the time I look.
It seems your experience is very different. Where should I be looking for some examples?
I’m confused by this comment. I follow the asahi Linux developments here on hn and read the comments. In this thread alone there are several high rated posts to the effect of “this project rocks!!”
The only reference I can find to trans- anything is basically this comment. Quite honestly, I recommend a thicker skin. Some people will be assholes, whether you’re trans, straight, bi, whatever. Pretending that assholes don’t exist will just make you angry… you can’t just wish away the assholes but you can ignore them and prove them wrong by being awesome (which in this case the asahi linux developers work speaks for itself!)
It seems HN's philosophy is like that of (early) Reddit, where the commenters self-govern and down/flag comments like that - and it seems to be working if they're all dead, right? They're hidden unless you have an account and specifically enable showdead.
> vile posts in the deads and a couple sprinkled inbetween. I’m not sure if HN staff doesn’t care about it, agrees with them or simply isn’t aware.
So basically you're judging a community by the posts the community and moderators deemed inappropriate and moderated away? Rather curious metric.
If you don't want to see dead posts then turn showdead off (which is the default). That you get some transparency in the moderation is a feature, IMHO, but it's also opt-in.
Maybe there should be a "verydead" for these kind of outright idiotic posts so they just won't display at all for anyone, but then how do you prevent abuse and keep the mod workload reasonable marking endless posts as "verydead"?
> there is constant transphobia against Asahi contributors on display on this site
Could you be so kind as to link me to an example?
People have discussed Marcan's anime alter ego, certainly, but I've never seen anything transphobic in the slightest. Maybe it's all flagged before I get to see it, but I honestly can't figure out what you might even be referring to. I enjoy reading Asahi updates when they come up, so I am unsure how I could be missing something so 'constant'.
> It’s not limited to the Asahi project, either. The site isn’t safe for queer folks.
As a 'queer person' myself, I find that statement utterly ludicrous and I reject it completely.
This is pure and poor conjecture, just like the rumor (originating on chan boards and KF) that a person close to marcan had faked suicide. Exactly this sort of rumor mill is what is wrong with HN linking to Asahi developers.
> As a 'queer person' myself, I find that statement utterly ludicrous and I reject it completely.
Are you a cis gay? I remember a few cases where people reached out to have information redacted to dang and it took weeks. While the people mentioned above were digging into private lives. This site absolutely requires you to shut the fuck up about your own life if you are at risk of being turned into a "lolcow" or pedojacked or whatever else these people will come up with. I understand that's not a consideration yet for 'queer people' currently not conscripted to the front of the culture war.
I can honestly say I've never, ever, heard anyone call anyone else these terms on Hacker News. I don't think I've ever even heard those terms before. It sounds like you might be very online, in a way that seems to be making you unhappy.
Some of my most upvoted comments on here have been from a 'queer' perspective. It helps to assume good intentions and engage with others constructively and in good faith.
I must admit, I feel saddened by your dismissiveness of certain 'queer people'. You have no idea who I am, what my identity is, and yet you so casually dismiss members of the marginalised group you purport to be defending. How callous of you.
Perhaps the reason I can so easily dismiss your hysterical claim that HN is unsafe is that - in my day - 'unsafe for queer people' meant 'reasonable likelihood of getting a brick to the face', and not 'seeing words online you don't agree with'.
this is very dismissive for how much you're projecting i'm dismissing you. it's okay if you can't relate to me, just don't assume you understand.
hn is unsafe in the way that it provides zero control over what you have submitted and a perfect history of your posts, mostly set in stone unless you email an administrator. at least reddit lets you self-service delete posts.
those are perfect conditions for a certain kind of group that spend way too much time digging up and correlating info to then start harassment campaigns that exceed 'seeing words online you don't agree with' quite often.
you can of course now go on to scold me and others with this problem about how we need to up our opsec or shouldn't post in the first place. i find such arguments, if you were to make them, entirely unconvincing. being aware of your risk profile is one thing, shifting all the blame for making it harder to retroactively rectify little pieces of information (these people found a place from a blurry 500x500 picture of a parking lot out a window) on the user is just a bad excuse for shitty UX.
> you can of course now go on to scold me and others with this problem about how we need to up our opsec or shouldn't post in the first place
Not at all. I neither think you should up your opsec nor avoid posting. I think you should be unabashed of what you have to say. I also think you should hear out others doing the same, generously and in good faith.
That tiny minority who harass and abuse queer people today? That used to be virtually everybody, all the time, everywhere you'd go.
The only reason queer rights are where they are today is because people weren't afraid to speak up, even when they had every right to be. When coming out meant admitting to criminal acts - proclaiming them in public, no less. Had they not put themselves out there - had they refused to speak to those that didn't already agree with them - this would be a much darker world, and your definition of 'unsafe' would be a lot more visceral.
People used to march, faces out in the open, in their small towns, for their rights, past neighbours who hated them. That's what 'unsafe for queer people' means. I'm just never going to be able to see 'people disagree with me online sometimes' as being in any way comparable.
A key tenet for many people appears to be that they should be able to do whatever is within their rights, but that leaves out the fact that sometimes doing something within your rights makes you an asshole.
> This raises interesting moral questions that I'm not sure I have an answer to.
It's not really all that much of a moral conundrum. Marcan's belief - expressed a number of times on his Mastodon - appears to be that he can prevent other people from discussing something, for the sole reason he doesn't want it discussed. It's not a particularly defensible position in an open society.
In particular, he is upset that people on Hacker News tend to point out that a contributor on Asahi - Lina - appears to be a computer-generated anime alter ego of Marcan himself.
Me, I have absolutely no problem with Marcan having an anime alter ego, but I don't think it's entirely reasonable to expect people to refrain from noticing this and remarking on it. Marcan disagrees, and this is the source of the HN-Marcan rift.
(As I've remarked before, I do mind OSS projects listing fake contributors, for both ethical and legal reasons, but that's another discussion.)
Quit with the gaslighting. It's not about an anime avatar, and you know it.
The problem is more that HN is perceived, with good reason, to be transphobic, and Asahi has several trans developers. This is also the reason why a couple of other people I won't mention don't want HN to link to their projects either, because for every link, several people will bitch in the comments about the validity of their gender and pronouns. (As if creating a web app somehow makes one an authority in biology.)
I mean, obviously, it's not like everybody who comments here takes part in the abuse, but when you spend every damn day of your life seeing it and depending on where you live, possibly being harassed by the government over it too, HN turns out to be one more place where you simply can't have any peace.
(And honestly, even if Lina is a anime alter ego, which you haven't proven, and only suspect, why the fuck do you care? Don't you have something better to spend your time on?)
I'm always so knowledgeable in the eyes of people who criticise me!
No, I didn't know Asahi had a single trans developer. I've never seen it come up on HN, where I most often hear about the cool work being done by the Asahi team. The vast majority of comments on here about every new Asahi article are effusive praise.
I dispute the characterisation of HN as some transphobic hellsite. That characterisation is simply not accurate, in the slightest, whatever perception may prevail on the Asahi Discord.
> why the fuck do you care
My other comment below, explaining the legal issue with fake OSS contributors, was already up by the time of your comment, so I refer you there. More broadly, I think Marcan should get to have as many anime alter-egos as he wishes. I'm just not particularly surprised that people find that noteworthy, and I think if you choose to have a public anime alter ego, you probably should be able to deal with that? I feel like Marcan's attempt to shut down this discussion is a perfect example of the Streisand effect - I certainly would have never found out about 'Lina' were it not for this silly feud.
> several people will bitch in the comments about the validity of their gender and pronouns.
I've seen these posts, and they're horrible, mean-spirited, and hurtful.
But as far as I've seen all those posts have also been downvoted to hell, flagged, and hidden, and the users often banned, and they're often from new "green" users rather than regulars. It's entirely possible that some of these posts remained (especially if they're posted after the conversation died down and there are less eyes), but I'm certain that if you email dang that he will take action.
At some point the first comment on one of my articles was "gay n---r soiboy" or something to that effect from a new account – a curious comment since my website has a picture of me being white enough to get a sunburn in Ireland – but how do you prevent that? Limiting sign-ups is the only thing I can think of.
I can definitely understand that people feel very negative about these things, but I think it's unfair to judge all of HN by it – you're essentially judging a community by the posts that were considered inappropriate and were removed.
And if you don't like HN (for any reason) then that's fine, so don't visit HN then. All this tomfoolery with referer blocks seems rather at odds with how the internet is supposed to work; "microsoft.com blocks links of the referer is from lwn.net" would cause a loud uproar here.
You are misunderstanding. You can discuss your conspiracies anywhere you want, just not on a post pointing to the people that have to deal with the fallout of your unfounded conjecture.
Also assuming there is a "fake contributor", who cares under which names contributions are split up? The work still got one. Also, it is absolutely not your project, you don't get to demand people show their ID when they write code for a project.
> Also assuming there is a "fake contributor", who cares under which names contributions are split up? The work still got one. Also, it is absolutely not your project, you don't get to demand people show their ID when they write code for a project.
If I contribute my code to an open source project, then I - as the copyright holder - agree to license my work under an open licence.
If I use an OSS project, I am using other people's copyrighted work under an open licence from them. Without that licence, I have no legal basis on which to use that work.
Only real people can (currently) hold copyright. If person X writes some code, but the licence (incorrectly) attributes the copyright to person Y, and person Y purports to give me an open licence to use that work, then - crucially - I have no license from the actual copyright holder (person X) to use their work.
Until an effective open source license is made, this code is not open source; it is completely proprietary. If person X chooses to sue you for copyright infringement, it is no defence to say that you're using it under a license from person Y, because person Y had no right to give you that licence.
This is a major ethical and legal problem. I would be very wary of the Asahi codebase.
If you wanted to split hairs this thin, you wouldn't use any project with at least one german citizen as contributor, since they can never truly yield all copyright on a work. You'll be fine. Anonymous contributions to free software (or even entire releases done anonymously, e.g. Bitcoin) are not actually uncommon.
As I said, very weird hangup to have, definitely not motivated by other reasons.
> If you wanted to split hairs this thin, you wouldn't use any project with at least one german citizen as contributor, since they can never truly yield all copyright on a work.
Nor can anyone else, copyright is not generally 'destructible'. That's why it's a licence. The holder keeps the copyright, but licenses the work to the general public. (Assignments are another way to achieve something like this, provided the assignee then licenses the work.)
I assume what you're referring to is inalienable moral rights - hence the reference to Germany - but those are a feature of many (most?) of the world's legal systems. They are included in one of the revisions to the Berne Convention, if I recall correctly, which is an international treaty on intellectual property.
I understand you're sceptical, but the legal dimension of OSS does matter. Using copyrighted material without a licence would constitute a major business risk. I would appreciate it if you could kindly refrain from making ungenerous assumptions about the intentions of others.
The most commonly cited example is a musician objecting to use of their music during a neo nazi rally. They won that case as the court judged the integrity of the work to be compromised.
I see. Well, I don't think that would affect code in the same way. If they have already contributed it then it's out in the open, and unless the maintainer was doing something objectionable they wouldn't have a case, and even if they did it would only be enforceable in Germany.
> that he can prevent other people from discussing something
And again, when some rando says "they want to prevent people from discussing, free speech!" the topic is always the same, they want to be freely racist, homophobic, whatever. Cmon man, just stop.
Forced participation has no place in open society. In such a society, when in private spaces, the valid response when told "You are not welcome here" is to leave and not to harangue the host about open societies.
Adding noreferer tag is “tasteless”? What’s tasteless is jwz for example silently redirecting visits from this site to goatse. This just reminds me of the clueless businesses in the early days of the web asking for “license agreements” to link to their content.
Its been more than a year I'm running asahi on my macbook air and I can't stress how grateful I feel for enjoying such wonderful freedom.
I don't feel like ever going back to x86 to be honest, at this point there is nothing lacking or unable to run and when the neural engine drivers come online now that the GPU is starting to mature people will be able to juice out every last bit of computation this machine is capable of.
For the record, I've switched to the edge branch a couple of months ago and honestly I noticed no actual difference in my day-to-day tasks which is really telling about how powerful even the M1 is when it can handle software rendering in such an effortless manner coupled with anything else running.
> at this point there is nothing lacking or unable to run
Sure there is. You just haven't run into it yourself.
Faster, cooler and more power efficient hardware is great. I just don't think that it makes up for depending on a small team of volunteers to resolve all hardware issues in an ecosystem hostile to OSS, which might break at any point Apple decides to do so.
And the incompatibilities with ARM are not negligible. If all your software runs on it, great. If not, good luck depending on yet another translation layer.
I'm sticking with my slow, hot and power-hungry x86 machines with worse build quality for the foreseeable future. The new AMD mobile chips are certainly in the ballpark of what Apple silicon can do, so I won't be missing much.
> depending on a small team of volunteers to resolve all hardware issues in an ecosystem hostile to OSS, which might break at any point Apple decides to do so
You are describing how most OSS software has been developed. I don't see how this is any different than early linux when no hardware manufacturers had any interest in supporting it.
A lot of the work that the asahi team is doing is just fixing Arm issues in the linux kernel (and sadly user space). That work will benefit everyone using Arm systems, not just folks running asahi on Apple hardware.
Its good for there to be more hardware architecture competition! I'm glad I can run my server workloads on the Arm servers in AWS that are 20% cheaper than the equivalent x86 machines. I'm glad that I can run the software I like (linux) on legitimately nice hardware (m2 air). You can make different decisions on what architectures are best suited for your needs, but the competition in the market improves the options and prices for everyone.
I've been using Asahi since the fall of 2022. When I first started using it a lot of software was broken because of bugs in that software that had never been exposed before (specifically around page sizes larger than 4k). All of that software has now been fixed. Support for linux/arm will only continue to improve as more people use it.
> I don't see how this is any different than early linux when no hardware manufacturers had any interest in supporting it.
It's very different. Hardware manufacturers in the 90s were incentivized to support Linux to expand their customer base. A trillion-dollar corporation has no incentive to sell their hardware to a niche of a niche of technical users who are not part of their software ecosystem.
Another major difference is that Asahi is a small team of dedicated volunteers who want to run Linux on their Macs. They're a niche intersection of Linux hackers and Apple fans. Unsupported hardware in the 90s typically had a much larger customer base and group of hackers willing to spend time adding supporting for it.
Even worse: Apple can decide at any point to make their hardware much more difficult to support. Newer models or firmware updates might break things. Being at the whims of a corporation that is the antithesis of F/LOSS to run Linux on their hardware doesn't inspire confidence.
> Its good for there to be more hardware architecture competition!
> Support for linux/arm will only continue to improve as more people use it.
Agreed. I'm glad that Asahi exists. But we've had ARM Linux distros for decades now. What Asahi is doing is specifically to support Apple hardware. Some improvements will trickle out to improve general ARM support, but this points out the gargantuan task they're actually up against. Not only do they need to reverse engineer the hardware, they have to resolve all software issues with Linux and ARM. My hat's off to them. The willpower, patience and skills required to wade through the absolute mountain of issues must be astronomical. Yet this is also part of my concern; how long can a developer keep the motivation and sanity to swim against the current?
It's great that Asahi works for you and everyone else. I'm just pointing out why it will likely never be my choice for any serious work.
> Hardware manufacturers in the 90s were incentivized to support Linux to expand their customer base.
That is not an accurate description of linux support by hardware manufactures from that time period.
> Unsupported hardware in the 90s typically had a much larger customer base and group of hackers willing to spend time adding supporting for it.
I also don't think this is generally correct. Have you looked at all the random drivers in the linux kernel for niche hardware. A ton of that is from one or two hobbyists taking the time to add support.
> Apple can decide at any point to make their hardware much more difficult to support. Newer models or firmware updates might break things. Being at the whims of a corporation that is the antithesis of F/LOSS to run Linux on their hardware doesn't inspire confidence.
I guess, but so what? Apple can't break the hardware they are already shipping if you are just running linux on it. Its true, I might not buy a theoretical future laptop from Apple if I can't run linux on it, but I don't see how that would affect my purchasing decision for hardware that is currently available.
> The willpower, patience and skills required to wade through the absolute mountain of issues must be astronomical. Yet this is also part of my concern; how long can a developer keep the motivation and sanity to swim against the current?
Hmm, maybe you've not worked on projects like this, or are motivated by different things. To me, reverse engineering a thing to figure out how it works and then writing software to get it to do things the original designers hadn't planned for is one of the more satisfying and fun activities of being a software engineer. I suspect the asahi team is having fun doing a lot of this work. (That's not to say its all fun. It sounds like getting things upstreamed has been trying. I also think having to read giant comment threads where people are needlessly negative about their work might be a bit demoralizing.)
> It's great that Asahi works for you and everyone else. I'm just pointing out why it will likely never be my choice for any serious work.
> Have you looked at all the random drivers in the linux kernel for niche hardware. A ton of that is from one or two hobbyists taking the time to add support.
Sure, for _niche_ hardware. When was the last time a GPU driver was added by reverse engineering it? The single Nouveau maintainer was burnt out, last I heard, and the project was never a serious alternative to NVIDIA's closed driver. Kudos to whoever found the energy to contribute to it, but these projects usually don't have a bright future.
Now expand that to include maintaining all Apple devices, and it's an insane amount of effort realistically unsustainable for any group of volunteers. But good luck to the Asahi team.
> Nouveau maintainer was burnt out, last I heard, and the project was never a serious alternative to NVIDIA's closed driver
Simply not true. I recall in the mid '10s using it because the proprietary driver was crashy garbage. No, I didn't get the same performance possible with the proprietary driver, and I didn't have a bleeding-edge video card, but it was more than usable as a daily driver.
>When was the last time a GPU driver was added by reverse engineering it
Freedreno (for the Adreno family)
Etnaviv (for the Vivante family)
Panfrost / Bifrost (for Mali)
All these RE efforts built on each other, although the GPUs are different the tools built to do the RE were shared (and I think ashai is benefiting too).
AFAIK Google has now hired Rob Clark the Freedreno maintainer who started all this to work on Freedreno for Android / ChromeOS
Upstream Linux now has pretty good GPU support for all the major mobile GPUs these days. The hold out has been PowerVR but they are now working on an official (not reverse engineered) driver.
Nvidia/Nouveau is a good example. I've had a number of laptops with nvidia graphics. For most of that time Nouveau was _more stable_ than the official nvidia drivers. Linux clearly was a second class citizen for nvidia for most of the last 20 years. Maybe go back and rewatch linus' rant about nvidia if you need a reminder about how terrible they have been historically.
Nvidia now only sort of cares about linux because of gpgpu applications. They still clearly don't care about gaming on linux; or desktop stability.
Yes, I will take Nouveau over the official drivers whenever I can.
In the same vein I used to consider intel drivers to be the most stable and best for desktop use. Now understand there are many caveats to this statement.
I use a lot of openbsd and you won't be running nvidea drivers on openbsd for love or money. there is nouveau and they are doing amazing work, however, they are also up against a petty, secretive company that appears to hate them. So nvidea is out.
Understandably openbsd gets zero support from the manufactures. So we need an opensource driver and some brave heroic soul to volunteer their time to get it running.
AMD drivers worked well enough and if you want decent 3d acceleration the only real choice. However they tended to crash and 3d acceleration is usually not a priority if using openbsd. Also starting with amdgpu the drivers got big, really big. The amdgpu driver nearly has more code then the rest of the openbsd kernel[1]. it is this big mess of generated code where each card uses a slightly different ISA. I understand why having a stable ISA is not a priority for AMD(it lets them change the card architecture easier) however sometimes i wish it were documented and pinned down. it would certainly make for a more stable driver that is easier to integrate.
And then there is intel, Note that I have not used intel graphics since 2016 so my experience is out of date. but once you got past the first generation of intel graphics the experience was rock solid, the drivers always worked well for me. if asked for the best openbsd experience I would recommend intel every time. However my last few machines have been amd and unfortunately there is no intel graphics add-in card(i looked). My last intel box I had an amd 3d card but I only used it under windows to play games. for work/openbsd I would just use the onboard intel graphics as they were more stable.
Nouveau has never been more stable, or nearly as performant as official NVIDIA drivers. I've had the exact opposite experience from you on every laptop I've had since Nouveau was released, so we must live in different universes.
> Linux clearly was a second class citizen for nvidia
And Linux is not even on the radar for Apple. :)
Anyway, I think we've exhausted our arguments here, and are just talking past each other now. Have a good day.
On some of the older devices, nouveau actually works very well right now. On these devices the open driver is as stable and performant as the closed driver and has all of the same features with the exception being GPU compute stuff.
I think mostly people misunderstand the limitations of the nouveau drivers and just assume they are bad in general, but it's very much dependent on the device.
One of the big issues right now is that nouveau can't bring the newer GPUs out of "idle mode". Nvidia has explicitly restricted this feature and the chances of the issue being resolved without cooperation from Nvidia is very low. I think a lot of people try nouveau on the effected GPUs and have horrible performance and then assume it's because nouveau is bad.
I have had an interesting experience with nouveau on the GT 710. For a little while there was a bug that would cause sway to crash back to tty, at first it would happen maybe once a day, then it became so bad that it would crash as soon as sway was launched. Now in kernel 6.3.3 it seems to be working flawlessly, which is how it was at some point in the past too.
Right now nouveau is working great for me though, so it is possible for nouveau to compete with the closed drivers in specific cases, but in general, for more modern GPUs it will have very low performance due to the reclocking thing.
Nouveau is an awesome project, but for later cards it's basically a dead project. You can get some features to work, but without proper power management there's no justifiable reason to daily-drive it. The proprietary driver is by no means perfect (particularly for Wayland) but it's the only real option if you own a modern card.
> And Linux is not even on the radar for Apple. :)
They must be awfully curious about why Xserve failed, then.
Claiming that Nouveau is or ever was as stable as the proprietary nvidia drivers feels like gaslighting to me, and I think that's what GP was saying. Nouveau has always been flaky.
If you've had good experiences with the nvidia drivers, thats great. My experiences with them have been bad and I will use Nouveau for general desktop environments unless I'm also doing gpgpu.
> the project was never a serious alternative to NVIDIA's closed driver
Because it can't properly clock the card since nvidia uses DRM to lock it out. Were it not for that, I'd be using nouveau every single day. In my experience it's more stable than the proprietary driver.
> reverse engineering a thing to figure out how it works and then writing software to get it to do things the original designers hadn't planned for is one of the more satisfying and fun activities
I feel the same way. I made a small driver for my laptop's keyboard lights and it was one of the most fun projects I've ever made. I can't even imagine how satisfying it must have been for the Asahi developers to get OpenGL running on that hardware.
> Hardware manufacturers in the 90s were incentivized to support Linux to expand their customer base.
Outside of some server niches (though even then, mostly not til the noughties), not really. You generally weren't looking at much first-party support at all in the 90s.
> Unsupported hardware in the 90s typically had a much larger customer base and group of hackers willing to spend time adding supporting for it.
This... is really not the case. The M1/2 Macs are _close_ to being one platform with something on the order of a hundred million units out there. Hardware in the 90s was pretty diverse; if you wanted support for your video card, say, there might be twenty chipsets out there, with maybe fifty manufacturers. Later on, there was a fair bit of consolidation, with virtually every computer, say, ending up with pretty much the same sound card chipset, but not in the 90s.
> Apple can decide at any point to make their hardware much more difficult to support. Newer models or firmware updates might break things.
Yeah, again, that _very_ much happened in the 90s.
> You generally weren't looking at much first-party support at all in the 90s.
I wasn't making a historical claim, but a philosophical one. Hardware manufacturers in the 90s had that incentive, if they wanted to pursue it. To Apple, a trillion dollar corporation, selling a negligible amount of laptops to a niche crowd of tech enthusiasts who will never become part of their software ecosystem is just not worth the effort to even pursue.
> This... is really not the case. The M1/2 Macs are _close_ to being one platform with something on the order of a hundred million units out there.
Right. And how many of those buyers are also Linux hackers willing to dedicate their time to work on projects like Asahi? How many of those would even be willing to run Linux instead of macOS? We're talking about an extremely small community of users compared even to the small Linux community in the 90s.
> Yeah, again, that _very_ much happened in the 90s.
Never said it didn't. Except that Apple is known for locking down their products, so the good faith they're showing now with leaving Macs relatively open can disappear at any moment.
> Even worse: Apple can decide at any point to make their hardware much more difficult to support. Newer models or firmware updates might break things. Being at the whims of a corporation that is the antithesis of F/LOSS to run Linux on their hardware doesn't inspire confidence.
That is the case with every single PC maker who doesn't ship all their PCs with Linux.
I just don't understand why Apple has to be treated differently. They just went through a traumatic transition where they could have locked their computers very tightly, preventing you from booting anything but macOS. They did no such thing. They did the reverse. No they didn't formally support it, but they own their own OS. How could they be expected to?
>A lot of the work that the asahi team is doing is just fixing Arm issues in the linux kernel (and sadly user space)
While I don't have Apple hardware so haven't been closely following Asahi I dont't think that is true. Linux has supported Arm for years (more like decades) now. They've been doing excellent work on support for Apple specific hardware sure, generic Arm not so much since it was mostly done.
Let me be more specific. There were a lot of bugs specifically related to non-4k page size architectures. Arm doesn't dictate the page size so most of the Arm systems out there have defaulted to 4k pages. The Asahi wiki has a partial list of userspace programs that were (or still are) buggy and broken because they made simplifying assumptions about how different architectures work[0].
Maintainers of other non-x86 architecture have said that this is improving things for them[1].
If you follow Hector Martin the lead Asahi dev he's encountered a number of bugs and race conditions in ARM linux systems which were never previously exposed because there weren't blazing fast ARM chips out there that could trigger them.
Why does Apple need to contribute to this work to make it somehow legitimate or good? I own some nice hardware (an m2-air), I want to run Linux on it. Asahi allows me to do that! Why can we not celebrate that the asahi team is bringing oss to new hardware?
They don't need to. I keep seeing that Apple does no less than other companies w/ regard to Linux. Well- where are their kernel contributions then? Lenovo and Dell (my two laptop manufacturers) contribute.
Cool, thats a different claim than what you said above but at least one I can actually engage with.
I've run linux on many Dell and Lenovo systems over the last 25 years. Most of those systems were fully unsupported by the manufactures for anything but windows. And yet, random people on the internet contributed to make that hardware (mostly) work. I've not seen any particular improvement in the driver situation since Dell started selling linux certified systems.
Its not really surprising though, Dell is just an integrator. All they do for their systems with linux pre-installed is to pick hardware that already has drivers. They took a little bit of work out of needing to research if a given configuration is likely to work or not with linux (which is good). They don't really deserve much credit beyond that.
Its also a little funny because most drivers from hardware manufacturers suck. I don't know why, but most hardware companies are terrible at writing software. Its easy to list off hardware companies that have a long history of shipping mediocre, buggy linux drivers: nvidia, amd, broadcom, realtek, (maybe i should just list every nic and wireless chipset manufacturer). Some of these companies have gotten better and have learned how to be good kernel contributors, but they were mostly bad for years and years. Thankfully in some of those cases random people on the internet reverse engineered the hardware and contributed from scratch drivers to the kernel. Most of the time I've been happier with the experience of running those from scratch drivers than what hardware manufactures ship.
Absolutely. A large part of the reason is that in the OSS world the architecture is optimised to make as much as possible common between drivers for different hardware.
For example for GPU drivers Mesa has tons of common code (NIR, GLSL parser etc) that is shared by all drivers with just the hardware specific parts being per driver whereas closed source vendor drivers reinvent the wheel each time.
Similarly for kernel wifi drivers there is a single MAC802.11 stack shared by all drivers.
Vendor drivers have an initial head start since those writing them have access to internal documents describing the hardware interface and don't have to do reverse engineering. But, over time, OSS drivers can be better as improvements to common code help all drivers.
In fact I think the best way hardware vendors could help OSS is not to provide drivers but documentation.
They also don't need to contribute to the Linux kernel. Why would they? They don't support Linux on their hardware in any official capacity, otherwise projects like Asahi wouldn't need to exist.
And playing devil's advocate, Apple has open sourced their macOS and iOS kernels, and has some open source presence[1]. None of their contributions are crucial parts of their business, of course.
I have an AMD Linux laptop I’ve been using for work.
It’s great. The battery life is great, it’s quite fast with a lot of cores, when I need to do my genetics runs (plugged in). Build quality isn’t bad, plus affordable and lots of ports. After my initial transition away, not missing my 2015 Mac book pro.
Linux is the way to go. I don’t blame people with apple hardware for wanting it. I just don’t feel the x86 side is as bad as the everyone makes it out to be. We’ve come along way since my first Linux laptop and it’s not so great battery life.
Two years ago or so I bought a ThinkPad with an AMD Ryzen CPU, there was a lot of hype about them. How Linux laptops were finally competitive, speed, driver, and battery-wise.
The machine was quite a bit slower than an M1 Air, would have loud fans during video meetings, and on Linux the battery would typically last 3 hours (6-7 on Windows, yes I did all the usual power optimizations). In S3 sleep it would discharge overnight and the next day it would refuse to charge with Lenovo’s included USB-C adapter. When waking up the machine from sleep the track point or trackpad wouldn’t come up 1/3rd of the time on Linux.
I used the laptop for work and the question ‘does the laptop work’ when having a meeting or having to teach became so stressful, that one day after another Linux hardware episode I immediately went to a store after work and picked up an M1 Air and never looked back (well, got an M1 Max after that).
There is no way I am going to touch Linux on laptops within 5 years.
(I use a headless Linux GPU machine daily, first used Linux in 1994, and was paid to work on a Linux distribution in the past.)
> Two years ago or so I bought a ThinkPad with an AMD Ryzen CPU
Things have changed a bit since then[1]. The new Phoenix chips are quite competitive with the M2 as far as performance and TDP goes. Your other complaints are with Lenovo, not AMD.
I doubt anyone will argue that Apple laptops don't have the best build quality. Apple has the advantage of full vertical integration, so it's very difficult for any other manufacturer to compete on things like battery life and power efficiency.
The Linux glitches you describe is the usual Linux jank. I don't disagree that even the most well-supported Linux laptop will have these. As a Linux user, you choose to deal with these issues because the alternative of relying on a corporation to decide how you're going to use your computer is not an option. I've also heard and experienced my share of issues with macOS and Windows. In the eternal words of a modern philosopher: every OS sucks[2].
Was the laptop advertised as supporting Linux, or, if not, did you at least do your research ahead of time to ensure that everything worked properly? Because clearly it didn't, so I expect Linux support was already known to not be in a great place before you bought it.
And yes, that sucks. We should have first-class support. It's no wonder macOS gained popularity among developers. But I've been running Linux on laptops for 15+ years now (even on Macs), and I've seen how it's changed from barely-working and having to futz with things at every kernel upgrade, to pretty much seamless (and these days I really have little patience for futzing around with things; I want something that works so I can do useful things on it). But, again: you need to choose your hardware carefully.
For reference, I had a 2016-model Razer Blade Stealth, which had no issues with Linux. Then in early 2019 I bought a 2018-model Dell XPS 13 that worked flawlessly (except for the fingerprint reader, which I knew ahead of time and accepted as ok). For the past yearish I've been using a Framework Laptop, which has had some problems (unrelated to Linux; Windows users have the same problems), but the hardware support on Linux has been solid.
Meanwhile, I'd constantly hear problems from my friends with Macs about how it could never stay connected to a wireless network after a couple hours (requiring a reboot to fix), or would frequently "beachball" under not that much load, or how the yearly major OS update would usually break their development setup. I used macOS on and off between 2005 and 2017 or so, and ran into plenty of issues as well.
While I certainly agree there's some laptop hardware that you just shouldn't run Linux on, the still-kicking Steve Jobs Reality Distortion Field somehow causes people to ignore or explain away all the issues macOS has.
Was the laptop advertised as supporting Linux, or, if not
Yes, it was Linux-certified.
I've seen how it's changed from barely-working and having to futz with things at every kernel upgrade, to pretty much seamless (and these days I really have little patience for futzing around with things; I want something that works so I can do useful things on it)
Dig up any post from 5, 10 or 15 years ago and Linux users will say literally the same thing.
Sorry it didn't work for you. I would recommend anyone buying at notebook to get one with linux pre-installed. I did that because I want to use this thing, not futz with it.
I'm assuming you're using Asahi Linux on your Macs (though you said you wouldn't touch it..). The lack of hardware diversity should make comparability easier, even if everything need to be reversed engineered.
I get 6 hours or so on my machine. Its pretty much silent, unless I push it. Its a Ryzen 7 5700u. We do a lot of parallel compute and genetics code tends massively parallel and x86 optimized. Mostly run on cluster though. I haven't done any maintenance and have had not hardware issues.
I don't link I could ever go back to macos, or windows.
What ryzen laptop is this that you keep referencing?
The negative experiences with the thinkpad are typical of all the intel laptops I have recently used, preloaded OS (including Windows, and to a lesser extent, Linux and MacOS), or not.
Whenever I look for an AMD laptop, it has a low resolution (1080p) display, and/or an off-center keyboard/trackpad (or has some other obvious fatal flaw).
I'm typing this on an M2 macbook. I do 100% of my work in an "8 core" arm Linux VM that can only use one core for userspace stuff (according to top), but that still kicks the pants off my previous laptop.
>Whenever I look for an AMD laptop, it has a low resolution (1080p) display, and/or an off-center keyboard/trackpad (or has some other obvious fatal flaw).
yup it has both of those. But the screen is only15", and I'm old so it doesn't matter. It not glossy which I really like though.
If you love the mac hardware, give Asahi a try. My understanding its the best linux for the M-series macbooks. Linux is great for developing on and they seem to be making great progress.
It's interesting how different tolerances are. I couldn't even stand 1080p on 13.3".
And then it needed some fractional scaling factor. Wayland apps worked ok with that scaling (though rendering was perceivably jankier), but some X11 apps would just be blurry. At the time there was no solution for there apps and looking at an extremely blurry CLion all day is no fun. The only solution was to run the whole desktop environment in 1x scaling and use this GNOME option to use larger fonts and widgets. Which worked ok-ish, but many things are sized in a funny way.
I just couldn't tolerate so much brokenness.
(And don't get me started on sharing a screen in Zoom conference calls.)
I don't understand how scaling is so broken. I switched my Linux desktop environment to a high DPI monitor in 2001 (it was a Sony CRT). Stuff like font and bitmap rendering looked better then than it does now.
The off-center keyboard thing is super irritating. I know some live and die by numpads but for my usage, 99% of the time they're just collecting dust and making typing less comfortable.
Laptop displays are also a common frustration, though this has been improving lately. Still too many models stuck on 16:9 aspect ratio though, which is suboptimal for anything but watching movies due to lack of vertical real estate. By the time you've factored in all the taskbars, titlebars, toolbars, menubars, tab bars, and status bars you've got a keyhole left to peer through. This is less of a problem for those using something like i3 or Sway where half of those bars are hidden but tiling WMs just aren't my thing.
I managed to get Ryzen 5850U/15W Thinkpad with 4k display in 2021.
There were some necessary firmware updates during the first year to get the suspend and power management working (afaik, not so many kernel bugs), luckily fwupd now works for Dell and Lenovo so fw updates are easy and there is a popup to recommend them when available.
For the past year it has been very stable running Fedora and thus using the latest kernels.
Now waiting for the Ryzen 7840U configured to 15W to be available with 64GB ram.
I would like to see the volunteer pool grow, but I suspect a lot of things in the Linux kernel are maintained by a small pool of volunteers--it's not like all volunteers (or paid developers) work across the whole kernel, after all.
Also, which are the virtuous hardware companies and what's stopping them from making compelling products? And since I'm already a Mac user, I've already supported this evil company, so what does it matter if I change out the operating system?
Not sure why you're fixated on the whole Linux kernel - we're talking about a small pool of volunteers supporting complex modules.
There are good laptops out there other than Macs... Lenovo, HP, Dell, etc all have offerings which are supported out of the box by Linux because they aren't using their own hardware or do contribute the necessary code to run FOSS OSs.
I'm not asking you to now dump your Mac. That'd be silly. Continue to use your Mac with Asahi, just don't complain if Apple decides to break anything at any time, and expect it. Know that what you type on now is already planned for obsolescence and most likely has intentional design flaws as shown again and again by people like Louis Rossmann.
When your Mac dies, I am asking you to not buy Apple for your next laptop. That's all that can be reasonably asked.
> There are good laptops out there other than Macs... Lenovo, HP, Dell, etc all have offerings which are supported out of the box by Linux because they aren't using their own hardware or do contribute the necessary code to run FOSS OSs.
For the love of all that is holy, name one model that has the following properties:
- 6 hours real life battery doing C++ development work.
- 7+ days suspend battery life
- 99.99% success rate resuming from suspend under linux (~ 1 kernel panic per year is OK)
- Centered keyboard and trackpad
- >> 1080p screen
- bluetooth, wifi, webcam, etc, etc, all work reliably (~ 1 device "need to reboot" failure across all categories, per year)
- un-noticable fan
- less than 10% permanent hardware failure rate per year
None of the last ~ $10K worth of intel machines I've used (including high end macs, linux and windows machines) met the above criteria.
I've managed to tick all your boxes on not one but two Lenovo ThinkPads. Ironically I've had trouble with suspend on my previous work Mac laptop running macOS. Also issues with WiFi on the same Mac laptop not present on other devices. Every single hardware failure within warranty I've had on both personal and work laptops have been Apple - LCD, battery, mainboard, dvd drive, just to name a few. Yet all the others I've owned have done their three year duty without fail. Very much a case of YMMV.
The end of the intel mac laptop line models were the worst laptops I've ever used. My last one went through 2 screens, the battery swelled, it couldn't charge fast enough to stay on while plugged in (and the CPU ran at 10% perf while charging, but only if plugged in on the left side).
Also, at any given time 6 of the keyboard buttons stuck, and the touchbar constantly phantom pressed the siri button when I pressed backspace. It got under 90 minutes of battery when brand new. It reliably kernel panicked on resume 1-2 times a week. It was also loud and hot.
Maybe I should give lenovo another try.
They completely screwed me over on a warranty repair right after the IBM acquisition. When I got the laptop back after over a month, it was diagnosed "no fault detected" and had a new symptom: it leaked high voltage from the backlight transformer into to the case during boot shocking the heck out of me!!!
Also, when I look at their web page, I always have the problem that they offer too many sub-configurations, and there isn't a button that says: "just give me the one that definitely runs Linux with the taint bit turned off, and only contains components that have had stable OSS drivers for over a year, and whose BOM hasn't changed for at least 12 months".
Ironically, Apple's web page is pretty close to having a de facto button like that.
My thinkpad does all those things... I expected really something crazier as a rebuttal. It sounds like you may have bought bad products and then bought a Mac when they had those specs and were happy then...
I wish more manufacturers worked harder to keep fan noise down to between none and barely audible without severely throttling the CPU and GPU or toasting your lap. My laptop shouldn't be spinning up its fan just because I plugged in something as pedestrian as a 2560x1440 60hz monitor (as my ThinkPad X1 Nano likes to).
Aren't lots of things in Linux dependent on a small team of volunteers? I know the Linux foundation owns the whole kernel, but in practice how many full time people work on the ext4 driver or whathaveyou?
/me glances at their full NixOS desktop that native builds for x86_64-linux and aarch64-linux. Heck, most all of it cross-compiles too. Native riscv64-linux needs more attention (upstream support is missing in places) but most of my config cross-compiles fine as well.
Every month that passes, with similar HN comments insisting it's a bad time... I wonder: "am I 'special' or just spoiled by nix(os/pks)". Or maybe, just maybe, people's expectations of their distros are shockingly low. And maybe rightly so at times.
The compute accelerator story on mainstream, non-patched Linux, with upstream software isn't that good at the moment. You're going to be waiting a while before you can do fun stuff like organize layers across the Neural Engine and GPU for ML models, something CoreML can do today. Compute using graphics APIs exists, but it isn't really the same and loses out on many features people practically want and are used to, and it moves forward much more quickly than graphics APIs e.g. Nvidia just released Heterogeneous Memory Management as stable in the open source GPU driver for x86. The Linux accelerator ecosystem in practice is just held together by Nvidia's effort, honestly.
We really need something like Mesa, but for compute accelerator APIs. I'm really hoping that IREE helps smooth out parts of the software stack and can fill in part of this, but the pieces aren't all put in place yet. You'll need the GPU for a substantial amount of accelerator work regardless of Neural Engine support.
I disagree that there is nothing lacking on these machines with Asahi, I still run into small nits all the time (from 16k page sizes biting back to software missing features). But my M2 Air is 100%, no-questions-asked usable as a daily driver and on-the-go hacking machine, it is fast as hell and quiet, it has nested virtualization and is the only modern ARM machine on the market, and I love it for that.
Is the Neural Engine/CoreML used in "normal" desktop apps on macOS, or is it limited to specialized ML centered apps? In other words, where should I expect performance improvements if there was a hypothetical Mesa for compute accelerators? Spontaneously, I can only think of image editors like Photoshop offering AI-based tools.
Has there been any ongoing work on this? It's been marked as "WIP" in GitHub for a while now, and I'd imagine it's one of the more complicated things to reverse-engineer.
For €2300 I assume you're looking at an M2 Pro model? Note that neither the M2 or M2 Pro Mac Mini currently have working display outputs[1] so no you should not. Apple changed the way the display outputs work in M2 so they're now dependant on Thunderbolt/DP alt mode support which is not implemented for any Apple silicon machine yet.
On the other hand a cheap M1 Mac Mini would make a great machine to try it out. The M1 Mac Mini is the best supported machine currently.
I don’t run Asahi on anything currently, but I do have two desktop Linux machines, an M1 Macbook, and have previously run Linux on an Intel Mac… I can see the argument for laptops based on battery life/heat/build quality, but for a desktop machine I’d need a lot of convincing to justify the price premium and risk of compatibility issues in choosing a Mac Mini over a SFF/USFF/Tiny desktop with fully supported hardware.
I’ve owned both windows and Apple computers, quite many of them, over the last 20 years. On average, the Apple machines last at least twice as long as the windows machines while still being fully usable. One could argue just based on that basic math that they are worth twice the price.
See, here's where the undefined nature of things comes in. "Comparably spec'd" needs to be conditioned on what task you're aiming for.
A "pure gold hammer" is a terrible idea and would also be terribly expensive. But asking for a "comparably spec'd" hammer presumes the absurd premise that the material of the hammer must be kept consistent regardless of its intended use just for the purpose of being comparable.
To preface, I totally understand the value proposition of Apple devices for some use cases, but it is important to recognize that they are aiming for certain workloads.
As examples:
I have one friend that runs virtualization workloads that require a lot of memory, a lot of storage, a lot of cores, but they don't really care about memory bandwidth, "having a display", or even the noise of the device. An older server with 192G of RAM, 24 cores and >8TB of storage can easily be had and upgraded within $1k, whereas a "comparable" Mac Pro costs upwards of $10k! (Of course nobody would use a Mac Pro for this workload, so the comparison is moot)
I also have friends that are digital artists. They care about having a high brightness and color accuracy display but otherwise don't do anything that taxes the computer. They also appreciate having high quality speakers and long battery life. Some of them run M1 Macbook Airs at the lowest 8G memory configuration for ~$800 (discounted new from other retailers) + a digitizer for ~$100, while the closest comparable non-Apple laptops are all premium devices upwards of $1.5k and even then they are still worse in the battery department!
As for myself, I do light dev work, virtualization, gaming, but also travel a lot and present at conferences. I use a GPD Win Max 2 for a little over $1k (early Indiegogo pricing). The closest Apple offering would be a 14" MBP, and configured as needed (32GB/2T) would be about $3800 and still be short a 4G modem and a couple of extra devices like a digitizer, game controller, and dongle for USB-A. ¯\_(ツ)_/¯ Can't win 'em all.
A micro ryzen 5600U build with really bad quality components can be half the price of a mac mini with equal CPU performance. If you upgrade all of the mac specs you can probably get a larger divide, but IMO maxed out macs don't make much sense for most people.
It's that focusing on one specific aspect of the system and compromising on everything else that produces the really big discrepancies. I tried to use PCPartPicker to spec out a rough equivalent of a maxed-out Mac Pro in terms of CPU cores, GPU performance, and RAM and SSD capacity, but still ended up at with at most a 3.5x disparity, and that's ignoring the GPU VRAM capacity limitation and features like Thunderbolt and 10GbE and assembly and warranty and support. If you want to assign $0 value to a large portion of a Mac's features then you can make it look wildly overpriced, but that's mostly an admission that it's the wrong product for your needs.
The years of the keyboard issues left a bad taste in my mouth, but I switched to a non mac laptop for my previous laptop and now I'm back again. The coupling of the OS and hardware really do make for a great user experience. I don't want to play games on my laptop, which is the only real use case where I hear valid complaints. I just need my dev environment and snappy research and communication.
A valid complaint from me is linux based container resource utilization. The only really good fix for that IMO is if apple did something like WSL2 or FreeBSD's linux ABI and had an efficient compatibility layer. For now I just run dev containers on my (linux) desktop.
Yeah, the M1 MBA was really so out of the norm in terms of value that it's pretty much impossible to find anything like it at the price point.
It literally turned me into an Apple person, essentially overnight. I no longer even switch on my windows machines.
There do not exist "equivalent"s to any Apple devices (I don't see them licensing the M1/2 chips to anyone else anytime soon). But depending on what you care about, a "comparable" Apple device could be 10x more expensive. Of course, for other tasks a "comparable" Apple device can also be _cheaper_ than any non-Apple device available!
Only looking at aiming for similar "longevity" (since the parent is using that as a benchmark), there are plenty of devices that have a useful life comparable to Apple devices at 1/4 - 1/10 the price.
There is one above in the server/homelab space. Items like memory and storage are charged huge markups* so if you need a lot of capacity of those you are going to quickly get into the 10x range!
As for longevity, if you consider software support ending as EoL, software/OS support for a huge swath of Intel iMacs (especially those with DGPUs) was dropped by Apple quite a few OS releases ago and you have to run community patches to keep them working. Whereas similar decade old hardware still run Win 10 and Linux out of the box.
*: Don't get me wrong though, the markups are for good reason. x86 platforms don't offer anything close to Apple's ARM chip memory bandwidth (which are closer to DGPU levels). Similarly, for flash/SSDs.
My M1 MacBook Air (8GB RAM) was cheaper than the 16GB Dell Latitude 13 i7 I bought a year and a half previously. I am much happier with my MBA, even at 8GB, than I ever was with my Dell.
You can get a used M1 mini for more or less 400€.
Get a glimpse of whats going on in your local facebook marketplace, most likely you'll come up with nice offers.
In my experience the experience for those is quite bad, as you're sharing that 8G across both the CPU and GPU...
Judging from the OP's post of 2.3k€, they're probably considering a maxed out version, which has a completely different experience since you can fully take advantage of the high memory bandwidth for hybrid tasks unlike the low-memory models where you're sharing the limited capacity.
Not the OP but I got a 13" and a larger model Mac of the x86 variety when they were still reasonably young and even though I eventually got all of the bits and pieces to work it usually pays off to wait until a somewhat larger distro supports the hardware as well. That way you benefit from a much larger crowd of testers and once they have no more issues you should be good to go.
Moneywise it was definitely worth it, both machines are still working many years later and have been pretty much trouble free after the initial bugs were ironed out.
If I was in the market for a new laptop right now I'd wait for a bit and then pull the trigger on the latest model with broad support.
I'd say getting a macbook or a macbook air would be worth it, but rather than spending that much on a mac mini I'd probably get one of those new ryzen mini-PCs like from Beelink or Minisforum. You could get something with a 7735HS 8-core chip, terabytes of diskspace and a shitload of LPDDR5 RAM for 500€ and it's as small as the mac mini.
New M2 mac minis start at $600 (8 core CPU; 8GB ram; 256GB SSD). You can probably find similarly specced x86 PCs for a comparable price, but this doesn't seem unreasonable. https://www.apple.com/shop/buy-mac/mac-mini
This is extremely gross 'Dang, and I'm wondering how you can possibly defend doing this. Seriously, why are you going to these lengths to evade a soft-block on links to specific sites that very clearly and explicitly are communicating that they don't want HN's traffic?
adding a config to your webserver to evade a soft block is directly enabling continued harassment. Its not passive, not "I'm just getting out of the way here", not "I can't stop this, not my fault", you are literally editing your webservers config file for a site-specific change to allow your forum visitors to evade an explicit block, a block enacted because of harassment.
I mean that's because you consider it harassment.
But if you consider it simple criticisms, suddenly its the person not accepting them, not harassment.
Celebrities deal with much worse, yet they don't consider "Taylor switft is ugly" harassment.
I wonder if the new Mac Pro's full PCI Express support resolves any limitations that prevents people from using GPUs over Thunderbolt on existing Apple Silicon hardware (this is apparently a hardware limitation).
Although the Mac Pro's PCIe extensibility makes it a pretty mystifying niche product from Apple without providing memory and GPU expandability, once Asahi Linux gets running on there you should be able to not get the full abilities of the latest Vulkan and full OpenGL 4.6 by putting in a recent AMD card. The open source Radeon drivers should "just work" on ARM as they do in the Talos II POWER-based workstation, if they can be stably initialized that is. Heck, Nvidia publishes a binary Linux aarch64 driver and they sound petty enough with Apple to try to make that work.
You could have Asahi Linux running and delegate any not-yet-supported hardware to the 7 PCIe devices it supports. Would be quite a mighty ARM Linux workstation. Again though - only if Apple has the PCI Express support for it.
It seems they didn't make any massive changes and instead just put switches on the existing PCI-E Lanes. That probably doesn't bode well for full GPU support :(
This is great work and I commend it. But in other threads people are acting like Asahi Linux hardware support is 100% complete. My fear is that if I were to go this route and purchase the hardware I'd be seeing fraction of the performance and capability I would in Mac OS. To be honest this blog post seems like the project has a long ways to go, not that it is nearly completion.
I just can't justify buying hardware from a company that is so hostile to developers and hackers as nice as it may be.
> I just can't justify buying hardware from a company that is so hostile to developers and hackers as nice as it may be.
I don't think it's hostile, I think they're just hands-off; they throw the hardware over the fence and say, "if you wanna make use of it, here's our software; if you don't like our software, sorry no docs but you're free to write your own". Which is exactly what's happening.
I mean it would be nice if Apple had released more documentation, but I totally understand if they don't want the burden of supporting it.
First, personally I don't care what hardware or software people use, if they are happy with the tools that they using then that's good.
That said, Apple has been very hostile to hackers over the years imo. Hardware being hard to repair, access, upgrade, etc. I think at one point they were making it virtually impossible to replace components because they were serial locked.
As far as I am aware, progress Apple as made has been in response to public image issues or changes in consumer laws within the EU.
Plus Apple software is heavily indebted to Open Source software so they very easily could be releasing drivers for their hardware instead of relying on community to do backwards engineering.
> Hardware being hard to repair, access, upgrade, etc. I think at one point they were making it virtually impossible to replace components because they were serial locked
You can only have so many flexibility in design with modern hardware — they are not fitting things into 5 cm “thin” chassics anymore. How exactly are such a thin device be repairable? Similarly to how old car motors could be tweaked with, you need special tools to touch anything in a modern engine. This is not against the customers, these are trade offs.
But even this way, apple devices have by far the longest lifetimes, macs, iphones will have 2-3 owners easily - so is it really fair to call them out, or is it just baseless emotional reaction?
Also, what you heard about locked down components resulted in better security, a much lower risk of theft, and a much more clean second-hand market (where you won’t be sold a phone with a cheap chinese shittier screen for example).
> But even this way, apple devices have by far the longest lifetimes, macs, iphones will have 2-3 owners easily - so is it really fair to call them out, or is it just baseless emotional reaction?
I would dispute this claim, e.g. Apple settles iPhone slowdown case for $500m[1], just the first link I found looking for "planned obsolescence apple" on DuckDuckGo. This is not exclusive to their iPhones as one can find with a quick search.
Which was about a very stupidly communicated, but genuinely good intentioned misshap. Iphones older devices with old batteries had cases where they no longer could provide sufficient voltage to the phone and thus it would reboot randomly for people. To fix the issue, they decreased CPU frequency so that it would no longer drain so high peak power.
But they should have made it an option (it is one now on these devices) instead, they might have even come out good from it (as an android manufacturer wouldn’t even care about such an old device at the time).
"But even this way, apple devices have by far the longest lifetimes, macs, iphones will have 2-3 owners easily - so is it really fair to call them out, or is it just baseless emotional reaction?"
Why reply that criticism of Apple must be purely an emotional one? Kind of diminishes your argument here.
"Also, what you heard about locked down components resulted in better security, a much lower risk of theft, and a much more clean second-hand market (where you won’t be sold a phone with a cheap chinese shittier screen for example)."
Apple could just release stuff that didn't break so easily too so no need to risk changing out a screen if it ain't broke. There are plenty of ways to increase security of the device without making it less consumer friendly.
Additionally since the context here is whether Apple has been hacker friendly or not, why shouldn't you be allowed to upgrade and change the hardware of YOUR device? As in, you want to put in more storage or change the screen to one that's better in some manner (maybe it's just cost) then you ought to be able to.
That is it should be the device owners choice whether or not to replace their screen with one from Apple or a cheaper one.
Going with the first link. The "best" phone is not globally available. The second "best" comes with three years of OS updates. I stopped reading there; three years is a much, much shorter lifespan than you get with iOS or MacOS.
Also, say you have one of these phones, and are in a major city, then break it. How will you get the parts you need to repair it? How many hours will you be without a phone?
With Apple phones, it's typically same day service to get it repaired. Worst case, you can get a new phone with your data mostly transferred, again, same day.
The Ars article you link is pointing out that Apple is dropping software support for laptops that are 6 years old. That's better than pretty much any other vendor.
As far as laptop repairs go, frame.work is probably the best non-apple option, but they don't have a fixed policy for how long replacement parts will be available. The story is similar for Apple:
says they provide parts for up to 7 years, and battery swaps for up to 10 (subject to part availability). I hope frame.work will be able to do better, but I challenge you to find any laptop company significantly better than this.
(Other than soldered ram and disk, I honestly don't care about third party parts. It's not like Apple replacement part markups are insane or there are significantly better parts available. I've definitely never used third party parts for other brands of laptops, even when they were available. However, I've been repeatedly screwed over getting other brands of laptops repaired, especially under warranty.)
Anyway, I get why Apple has a bad reputation for support and repairability. There objectively bad. However, that doesn't mean they're not simultaneously also the best option (or close to being the best).
None of the machines on that list are more hackable than a Mac though.
Some are more "hackable" than an iPhone, but only in some strange symbolic way, since once you de-googled android is somehow more user-hostile than iOS and completely unhackable in practice (in the sense that I can't make it do what I want and also be usable as a daily driver that lets me do stuff like pay for stuff, use public transit, charge my car, park, or take an uber/lyft).
I'm a happy owner of a pine book pro, and a pc engines router; I get it. However, I don't think there are any viable Linux laptops or phones that compare favorably to the linux laptop + phone I had a decade ago.
Scratches head at how framework laptop is less 'hackable' than a Mac but okay.
"(in the sense that I can't make it do what I want and also be usable as a daily driver that lets me do stuff like pay for stuff, use public transit, charge my car, park, or take an uber/lyft)."
So either the hardware you bought isn't letting you do what you want with it (this includes iPhone too!) Or you become a better hacker and get it to work yourself.
> I'm not arguing over better hardware (performance) or price
Neither was the comment you replied to.
They were arguing that Apple provides excellent hardware support (in terms of repairs and how long they take) and lifetime (in terms of how long updates are provided for the device).
Hackability is irrelevant. Apple clearly doesn’t target that market and hadn’t since the Apple II days.
Would you buy a Lamborghini and complain it makes a terrible truck?
That’s not something Apple designs their products for, and their users are either fine with that or willing to make the trade off for the other things Apple does prioritize.
So the only option has 1/10th the hardware of an older iphone with shittier camera than that.. for me that’s not a good deal. Framework laptops are insanely expensive for again, worse hardware. Isn’t the point of linux to run on anything?
Also, every device needs battery replacements, like this is just the physics/chemistry of batteries and it has an absolutely doable price for any apple device.
What breaks easily on an iphone? They are quite sturdy phones with metal casing. They wouldn’t get sold after 5-7 years of active use if they weren’t sturdy. And glass will still be breaking when it meets with big enough force - I again don’t see your point.
> re hacker friendliness
The RAM has different architecture on the M series, so it can’t be replaced even theoretically. Also, every moving part is one more point of potential breaking, plus it takes up space. This is not a rasppi, different design goals/constraints.
You can put in a worse screen but one will be able to see that in the settings so they can’t be scammed.
> Framework laptops are insanely expensive for again, worse hardware.
1199 EUR is not insanely expensive, specially considering that I can put up-to 64GB of RAM in a Framework laptop with a reasonable amount of money, while I would need to pay almost the price of a full Framework laptop to do the same in a Macbook Pro [1]. This is IMO, insanely expensive.
And yes, I can definitely use those amount of memory during mass rebuilds that I sometime like to do in NixOS. I don't even try to do those same workloads in my macOS because they start to become hugely slow once you hit the swap.
[1]: by the way, this get even worse considering that I also need to upgrade from an M2 Pro to M2 Max to have the option to do so. I just did a quote for the cheapest Macbook Pro with 64GB of RAM, and I got a 4000EUR quote for 512GB of storage that is laughable low for something that expensive. At that price, I can get 2 of the most expensive Framework AMD and I would still have sufficient money to get another one of the older Intel ones as a spare.
Sure, I never said so. It doesn't mean that there isn't workloads that benefit from more RAM (even if it is slower) or that the Apple's RAM prices are insane.
If you like the hardware or software that you use then I am happy for you.
"This is not a rasppi, different design goals/constraints."
That we can agree on.
Apple is a business with specific goals and so far as a business in terms of profits they have been successful.
All I wanted to point out was that Apple is not hacker friendly in my opinion, and I have listed good reasons that you don't want to accept. There's no amount of going in circles here that will change either point of view I fear.
> That said, Apple has been very hostile to hackers over the years imo. Hardware being hard to repair, access, upgrade, etc. I think at one point they were making it virtually impossible to replace components because they were serial locked.
That came partially out of the desire to reduce the lure for thieves and robbers. It was really bad during the first generations that regularly had jailbreaks and ways to bypass "Find My..." or whatever, then the first tightening reduced resale values of stolen iPhones by a good amount (as they were only good enough to slaughter for parts once reported stolen), and the latest round made it even worse for criminals.
Personally though, I'd preferred they simply provided "unlock codes" with a phone that could be used to remove the association between a part's SN and the IMEI/SN of the phone. That way, buyers of iPhone have something similar to a certificate of authenticity.
In fairness, most instances of them doing that actually significantly increase the cost of evil maid hardware tampering attacks.
If I could, I'd configure grub or whatever to serial-lock my Linux install to my desktop hardware (and keep a recovery key that would unlock it at another location).
The issue is that Apple isn’t giving anyone access to the tools to pair the parts, unless you give them all the information in advance, buy them at possibly inflated prices through their self repair program if they’re even available, and then have Apple remotely approve it afterwards (and this process only really works for individuals, 3rd party repair is more important as most people don’t have the skill)
If I was running a third-party unauthorized repair shop, then I'd care.
However, I just don't see why I should care, as a customer. Also, I have trouble wrapping my head around the standpoint that people want to pay the Apple "it always sort of works" tax, then spend lots of extra time trying to save $50 on a repair.
When I mess around with hobby hardware, it's esoteric stuff, not readily-available laptops that are being built by the millions. For that, sometimes you need a volt-ohm meter, or diagnostic rigs.
Every few years I even bring some piece of a Linux box to a repair shop, to narrow down some fault.
However, Macs are all identical, so repairs boil down to "yank component, throw in the recycling shredder box, and (if you just replaced the mainboard) restore from iCloud Backup". Apple hardware switched over to being livestock, not pets a long time ago, so I'm not seeing the point in spending lots of effort on custom repairs (vs. replacing + recycling the bad components).
Because as a customer you should have the choice to get a repair from whoever you want. Even if you only go to Apple, competition will force their prices down somewhat.
"pay the Apple"
It's not just Apple doing this
"then spend lots of extra time trying to save $50 on a repair."
What if you're trying to save precious data? Apple don't make much effort to do data recovery because they don't care; you can fund 3rd party repairers who absolutely do. What if you got the device second hand? What if it's not $50, but $1000+? I've seen a video where Apple tells a journalist they have to buy a new Mac, and a 3rd party repair shop fixes it for free because it was such a small issue that Apple didn't even check for. https://www.youtube.com/watch?v=o2_SZ4tfLns
"Macs are all identical, so repairs boil down"
This is the problem; Apple is pairing parts to devices, so if you use a 100% genuine part from a new device it still won't work
> What if you're trying to save precious data? Apple don't make much effort to do data recovery because they don't care; you can fund 3rd party repairers who absolutely do
I keep backups, both in iCloud and elsewhere. Apple makes it a headache to avoid doing this, and they apparently finally added proper E2E encryption support.
I guess I think of Apple computers more like generic corporate fleet vehicles, and Linux boxes more like classic cars.
I really don’t think anyone cares if a ford dealer irreparably destroys 1% of the bottom trim white F-150s they maintain, and also forbids the people that bought the service contract from opening the hood.
None of that applies to a mustang shelby, but that’s a different part of the market.
Maybe some people really like their macs / phones, and don’t just use them like interchangeable boring tools.
> If I could, I'd configure grub or whatever to serial-lock my Linux install to my desktop hardware (and keep a recovery key that would unlock it at another location).
This is the general idea behind TPM/Secure Boot, but as you present it, it just sounds like a headache for performing system recovery, at no obvious benefit for security.
What's your threat model? In the 99.(9)% case it's a crook snatching the laptop, wiping the HD, and selling the whole thing and/or the parts. Evil maid is a real threat, but only practical (in terms of sophistication/cost vs benefit) for high-value targets, like C-levels, devs holding company secret keys, etc.
I don't really care if the desktop gets snatched. I'm more worried about old tax returns, the credentials stored in its password manager, etc. Breaking in twice (once to install a bump in the wire key logger, and once to walk away with the machine) would lead to at least 10x more payout for a burglar.
Also, ignoring what it is worth to the attacker, having to roll over all my credentials, freeze accounts, etc, etc, because my desktop was stolen would cost way more of my time than buying a new desktop (happily, the drive is encrypted).
> I think at one point they were making it virtually impossible to replace components because they were serial locked.
That seems like an obvious reaction to the fact that criminals (once they were prevented by activation lock from selling stolen iPhones) started parting out stolen iPhones and selling the parts to repair shops.
I think a better approach would be for Apple to only block the replacement part if that part has a serial associated with an activation-locked phone, but I'm sure it's easier to just block everything except for the replacements shipped directly from Apple.
That's not true for Macs. Apple allowed and worked with MS to allow windows to work on Intel macs. That's pretty insane from allowing people to do what they want with their hardware.
You claim Apples Mac line of products is adversarial to hackers. Which I dispute with the example they help alternative OS to run on the Mac. First with Windows and now Asahi.
There's a line somewhere, and I think it's different for every person. For a lot of people, the pricing itself is just outright hostile. 8gb of RAM in a base model Mac is not future-proofed in an age of local AI models, and paying to upgrade it gets expensive, fast. For others the OS is hostile when it unceremoniously drops support for $THING they use, or because it has the gall to show them ads. For others yet, the hardware is hostile because it's basically a black box that Apple withholds documenting to maintain a monopoly on fixing them.
Apple is one of the few companies smart enough to deliberately do this. It is both a testament to ability to do brilliant things, and akin to being trapped in a room with a lion that has twice your SAT score. The "golden handcuffs", as they say.
There is literally no other laptop on the market that would be anywhere close to an M1 air in terms of performance-battery life combo. And it is definitely not even on the upper end in price for that, tiny windows laptops are ridiculously expensive!
Well overall if you sum up what even a basic m1 mb air packs, be it the panel, quality peripherals and soc itself its rather reasonable and within reach for almost anyone contemplating of buying a new computer. When you factor in that they were the first ones with an actual capable desktop arm offer I'd even say its cheap.
At least this is how I felt 2 years ago back when I bought it and more or less even to this day. I'm wondering how is the market as of now, two years later, and how the $1k arm laptops coming out today compare.
Macbooks are really cheap for what they are, not sure what are you on friend. I would never buy one though, still, they are very impressive and its a shame that they are not free.
> Regarding the capabilities not sure which one you miss. Do you plan to use it for development, or you want some kind of gaming/multimedia setup?
> It's the other way around. It has been usable as a daily driver for ages.
Honest questions since I haven't been paying attention to Asahi for some time now:
- Does hardware accelerated video decoding work? Including in Firefox?
- Does sleep work properly or do I get significant battery drain after leaving it sleeping during the night time? Also, does it wake up from sleep reliably? Like if you open/close the lid 100 times in a row would it crash?
- How is wifi? Does it work as fast and reliably on Linux as the Intel cards? Supports latest WiFi standard and 6ghz?
This would be my most basic questions to buy MacBook as a daily-driver Linux laptop.
I don't see why it is unreasonable to desire Apple to provide documentation and open up their hardware. Honestly I'd ask the same as any other hardware vendor. Intel for example provides very detailed technical documents on their new GPUs (A770).
> Intel for example provides very detailed technical documents on their new GPUs
Intel sells parts to other people, and those buyers expect that kind of documentation. And when it comes to GPUs they’re in 3rd place for high end GPUs and are incentivized to give people as much info as possible to optimize for their hardware.
Apple doesn’t sell their computer hardware to any other OEMs, or hsve a 3rd party making their OS, so they don’t need to produce documentation for anyone.
I agree it would be nice if it was documented. But absent a law forcing them I don’t see it happening.
You don't create a new bootloader that allows users the freedom to run an unsigned third party OS without having it degrade the system's security if and when they boot the native OS because you are "hostile to developers and hackers".
I could trick you into adding a hardware key, then install a tampered version of Windows with it.
(Also, the last time I looked, TPM keys could be grabbed with ~ $100 of hardware, but I think that's fixed by some newer standard.)
But, yeah, it's not a big tradeoff in practice. I think their point was that Apple had to expend effort to enable the use case, which isn't "hostile" toward the use case.
> I could trick you into adding a hardware key, then install a tampered version of Windows with it.
There are other issues as well.
For instance, on a PC the security settings are applied per machine and not per partition, so you can't mix an unsigned OS on one partition with full security on another partition.
Also:
> On Wednesday, researchers at security firm ESET presented a deep-dive analysis of the world’s first in-the-wild UEFI bootkit that bypasses Secure Boot on fully updated UEFI systems running fully updated versions of Windows 10 and 11.
Despite Microsoft releasing new patched software, the vulnerable signed binaries have yet to be added to the UEFI revocation list that flags boot files that should no longer be trusted.
> For instance, on a PC the security settings are applied per machine and not per partition, so you can't mix an unsigned OS on one partition with full security on another partition.
This wasn't true for a low-end Acer I bought a while ago, and it's not true on an Asus motherboard I use. You can add keys to the bios, and then it'll let you run with either key. That lets you use the grub shim key. On the Acer, you can even tell it to screw PKI, and just check that the hash of the bootloader hasn't changed.
The gaggle of moving parts that are involved in the PC world make security and privacy substantially more challenging because of nonsense like this - a vendor with rubbish security (not even an HSM for critical signing keys!) compromising the broader world.
> I'd be seeing fraction of the performance and capability
You'd temporarily lose some hardware support (documented) while it's being worked on. But I'm not sure why you expect losing performance? This is running native code. Same binary will run the same on both systems (+/- the llvm version differences).
Does the CPU run at similar frequencies between Mac OS and Linux (since they're writing their own drivers this isn't guaranteed)? Is the scheduling done similarly? Are there any special hardware modes you have to activate with e.g. binary blobs?
There are a bunch of factors that could affect performance even under the same OS (try underclocking your CPU or play around with schedulers). Given the mostly non-existent documentation from Apple I'd strongly suspect that average-case performance will stay worse on the Linux side for a long time.
The currently top-rated top-level HN comment goes into those details. This release significantly improves CPU power management, to the point where it should be similar to MacOS.
Some of this stuff is handled by binary blobs that get installed/upgraded by MacOS, and are running by the time Linux boots.
With the previous release, power per watt and absolute performance were already better than high-end x86 laptops, so if your question is "is this faster and more power efficient than my other Linux laptop?", the answer is probably yes.
If you're asking if it will beat MacOS's perf/watt in all scenarios, the answer will be no for a long time. However, it is probably already beating MacOS in many practical scenarios.
Right. Everything I’ve seen implies they’re both using the hardware the same way (no “limp mode” or such).
Linux tends to be faster because the kernel is just far better optimized than Apple’s appears to be. Massive sums of money (and person hours) have been spent speeding up Linux networking, file systems, scheduling, etc. and it was all sent back to mainline.
Apple’s kernel team can’t possibly compete based on resources alone. They do their best but MS, Google, Amazon, Redhat, and so many others are constantly improving Linux to squeeze out every last drop of performance.
So in many ways Linux is better optimized. I remember one of the developers posting a few weeks ago about just how much faster code compilation was under Linux, because the file system layer is so much better. It was like 6x or something. Sort of an accidental ideal benchmark for stressing that.
E.g. the performance you can get out of the GPU at the moment is a subset of what you can get out of the hardware. Or as another more generic example, until this latest release CPU boost states weren't enabled due to lack of proper cpuidle driver which resulted in regressed single thread performance.
There is nothing inherent about running Linux that will require it be slower, in some cases it will/is even faster, but the lack of everything being fully supported does actually impact performance right now. It has been getting better with time.
Specifically GPU drivers, which can dramatically impact performance. Especially if I am attempting to run any kind of ML workload from Linux. I'm assuming it's basically a non-starter at this point and one is forced to use Mac OS.
Yes. I put that in the capability rather than performance basket though. As in, you can't access the compute shaders yet, rather than: you can and they're slower.
> I just can't justify buying hardware from a company that is so hostile to developers and hackers as nice as it may be.
While I understand what you’re saying, the developers of Asahi have said before that Apple appears to be going out of their way to leave things open for them.
Apple’s boot security is enough they could easily prevent anyone from ever running a non-Apple blessed OS. But Apple made it fully supported to boot directly into a non-Apple OS as your primary OS.
They’re not helping by giving code or documentation, but they’re not putting up roadblocks. Apple seems to be happy with a stance of benevolent indifference.
It’s not official support. But a number of PC vendors don’t give official support either.
A car with one seat seems 100% complete if your use cases only involve you driving it alone. Asahi Linux is absolutely in that kind of spot right now. For some people there is 0 reason to wait, for others it's not even worth booting. If you have fear it's not fully complete enough, I'd say trust those feelings. At least right now.
There have been multiple reports of the last couple years that Apple has been informally internally helping the Asahi Linux team to make it run well on their hardware. Apple cannot come out and officially support another operating system of course, but they are aware of the interest and are helping make it happen, in an unofficial capacity.
Boot Camp was officially supported. It gave them an advantage because it helped people move over to a Mac from windows because they could still run some of their software.
If Apple was to help Asahi officially, I think they would be worried that they would be on the hook either for continuing to support it or for damage done to the hardware by other peoples changes to the code.
Basically I think it’s a combination of not wanting to spend the money and being worried about the potential downsides of doing it.
The upside, in terms of additional sales, is likely very small.
> I just can't justify buying hardware from a company that is so hostile to developers and hackers as nice as it may be.
As opposed to what company besides those tiny ones? Almost all of them are closed-source only and drivers have been painstakingly reversed engineered over decades.
Yeah, I'm not gonna run out and buy a new Macbook just to wipe it. But these M1 machines are going to be things you just have lying around sometime soon. If my wife upgrades, for example, I know exactly what I'm doing with her M1 Air. :D
Asahi is designed to be dual-booted right now (in fact you have to go off the beaten path to not do that), and it even uses the native Apple boot UI to let you pick your OS
Thanks to the entire Asahi team, your work is truly incredible and so far beyond my pay grade that words fail me. I honestly recently tried and very much struggled to communicate why I was so amazed by your project when talking with friends.
For anyone interested into the GPU side, I can't recommend Linas streams[1] enough.
Sample-rate shading is exceptionally rare (MSAA is rare-ish these days, but I only know of exactly one title that has shipped sample-rate shading), so requiring a basic compiler transform to handle it, especially when they can do so easily because of their tiler architecture, is pretty sane.
The work of the Asahi team is incredible, and so much fun to watch unfold.
I wander if now that you can get a rack mount Mac Pro with Apple Silicon (launched yesterday, the second coming of XServe), running server workloads on them with Asahi Linux becomes a viable route for some people?
If you're doing on-the-road visuals and you want a chunky realtime rackmount to render with, $7k for an M2 Ultra might look like a steal. For hosting purposes and parallel workloads though, there are many other options at much better price/performance points in the same market. If you're not leveraging Apple software and you pay for an Apple Silicon server, you're kinda wasting money that could have gone towards an Ampere or Grace instance instead.
Are you aware that Apple made it clear yesterday they have a full Direct3D 12 translation layer yesterday?
The biggest problem is chicken and the egg. There is no high-end gaming hardware, there are few games, and there isn’t much demand for the kind of games Windows gets.
But without the hardware it’s not clear the games makes sense. And without the gamers it’s not clear porting makes sense. And without the hardware the gamers and the games won’t come.
Even if Apple released a 100% functional and performant Vulcan implementation tomorrow I seriously doubt it would make any real difference to the gaming market on the Mac.
Sorry, but what's the point? Why not just buy a Linux laptop and have everything work out of the box? Why are Linux enthusiasts putting so much effort into supporting hardware from companies that - at best - ignore Linux? This question is also valid for other manufacturers, btw, not just Apple. So much time wasted doing free labor for hardware companies that will just break your stuff with the next hardware iteration.
For me, the Macbook Pro 14 with a 10-core M1 Pro I am typing on has my favorite keyboard and trackpad I've ever used on a laptop (2nd place was the 2013 MBP and this feels the same). It has the best display I've ever set eyes on.
The battery life is the best of any laptop I have ever used by far.
And the performance for number crunching is as high as any x86 machine I had previously and per watt it blows everything out of the water. And it is dead silent while doing so, whereas every x86 'work' powered laptop I ever used would wind up sounding like a jet engine with my workloads.
So for someone who runs linux.. if they want to run it on hardware that is this nice (to me at least)... this is it. This is worth it. It's worth developing for.
Also, Apple is going to stick with the M-series SOC's for a long time now that they have switched. And they tend to keep hardware interfaces for a long time too. So the development of Asahi now will bear fruit for... at least the next decade I'd say.
I still use OSX for daily activities, but the kernel Asahi is developing may be my plan to stretch this 2021 M1 MBP 14 out hopefully to the year 2030, as MacOS moves on. My 2013 intel i7 Macbook Pro made it to 2021... 8 years of daily use and world travel. I was beyond the moon with that product performance and I'm expecting similar from these new macbooks based on my current 1.5 years of use.
> Why not just buy a Linux laptop and have everything work out of the box? Why are Linux enthusiasts putting so much effort into supporting hardware from companies that - at best - ignore Linux?
Almost all laptops sold with Linux pre-installed or with support advertized only work well with Linux due to volunteer work similar to what is being done with Apple's stuff right now. Almost everything is proprietary with close to zero upstream support, you just don't notice it because the work has already been done.
In an OSS ecosystem, manufacturers inherently have to use work that has been done by volunteers! The real question is whether they invest their own resources too. For example, System76 has created their own distro, Pop OS. My team and I were happy with the laptops purchased from them and the OS integration provided a smooth experience.
Are there commercially available arm-based linux laptops? I'm not after horsepower just decent battery life and enough computing power to do non-AI programming.
The point? There are no other ARM laptops that are even close to being competitive right now. I want a good fanless laptop. My only choice is Apple. Asahi made it possible and I now have a great experience. What's the problem?
Please tell me where is it because I will buy it instantly, and I’m only half kidding.
I don’t see how making linux available on possibly the currently available best laptop hardware any different to the previous decades of hacking a working wifi driver into the kernel. It was always an uphill battle, and we should be thankful for those who take up the hard work!
The only exception is the Bonobo, but it comes with a discrete GPU, so its battery life + weight are going to suck. Also, its keyboard is off center.
Most of the star labs machines have low resolution displays, but I can find nothing wrong with this one. If you choose AMD, a reasonable config is $2600, which is comparable to Apple. However, they are only building 10,000 units, and taking preorders, so it might be unobtainable:
The purism offering seems OK except that it is a 10th generation intel, and those were extremely bad, even by recent intel standards. Maybe they'll get an AMD refresh out the door with the same ergonomics.
The HP has an off-centered keyboard and trackpad and a 1080p display.
So, of those four vendors, there's one model that's vaguely competitive, but it's a limited production run pre-order.
Any? https://wiki.postmarketos.org/wiki/Lenovo_IdeaPad_Flex_3_Chr... is dirt cheap and runs PMOS quite well. Actually Chromebooks in general are a good bet when combined with PMOS - or even not combined with PMOS, if you can work with a normal Linux VM on top of ChromeOS.
You said "ANY arm processor". If you'd like to further constrain the exact requirements, that would be a useful clarification. (Although, I suspect that some Chromebook will still work)
Apple makes good hardware and makes good engineering decisions. Sure other laptop makers have low end and high end options. But frustratingly often match apple on 50-75% of the features. Sure some have nice still aluminum bodies. Some have screens that match or beat apple's. Some have nice centered quality touchpads and keyboards. Some have great battery life. Some have great performance.
Very few match on all aspects. I'd tried a few and always had one terrible issue. Terrible battery life, lousy screen, and/or terrible touchpad. Apple does seem willing to make improvements without as much worry about backward compatibility. There are a few that match on everything I care about, but often cost more than the Apple.
People like to complain and mention byzantine purchase methods. Wait for a lenovo sale, buy the bare bones model, apply the discount code, then buy dimms and SSDs from random bargain basement sellers. Oh and buy the linux compatible wifi card and do surgery on your laptop to get wifi working after suspend.
MBA is pretty compelling mix of performance, size, cost, and battery life. Unlike any x86-64 laptop, you can pay $500 more or so and get double the memory bandwidth. Or another $500+ and double it again. Definitely makes the macs better any PC at some workloads. Sure some x86-64 with a nice discrete GPU is way faster ... when plugged in to wall power.
They don't exist, or if they do, they rely on far less (motivated, whatever that motivation might be) community folks to bring-up support. Which often happens are the device is a generation or two old.
Non-sbc aarch64 devices with good upstream support are damn near non-existent. I understand the Linux user market is small, but it's baffling to me that companies can't invest a small amount in having coordinated upstreaming efforts.
That said, I'm so delighted with my G14 2022 experience that I might just buy the 2024 version of it's all AMD again. I have better hardware control in Linux than Windows, I can play Halo Infinite in Linux (praise be Proton) and Windows. But... I suspect I'm going to be able to do the same thing on a likely more-battery efficient and nicer hardware M2 MBA by the end of the year. Nice to have options.
Edit: it's really soulcrushing to think about. Sdm845 devices are insanely cheap and yet upstream support still barely limps along, mostly with volunteer/donated efforts.
> Why not just buy a Linux laptop and have everything work out of the box?
Because these things are rare as gold in the first place as being Linux ready isn't a focus for most OEMs, sometimes severely lag behind the competition in feature support (e.g. limited to UVC webcams with crap quality), have serious availability issues (Framework), you have a tough time getting service or spare parts, or barely any resale value, or limited choices in screens (which is the one and only thing keeping me from a Framework - who the fuck wants a 3:2 screen?). Also, tough luck getting firmware updates for embedded components.
Apple devices, spare parts and repair centers, in contrast, are widely available across the world (okay, maybe not in places sanctioned by the US), firmware updates come around when needed and hold their resale value for years.
Are there any PC laptops, running Linux or Windows, that you can buy for a similar price that run as long, as cool (temp), as fast, and with a screen as good as an M2 Air?
I know there are faster machines, but they’re usually hotter. There are cooler machines, but they’re usually not as fast. There are cheaper machines but the screens are worse. And there are long battery life machines but they’re usually low performance Chromebooks.
I don’t know of any other machine that made the trade-offs to end up offering a similar combination. At any price.
And I said nothing about noise or thinness.
Once you’ve used one, it’s a very seductive combination of features. I don’t want to go back to “6-8 hours of battery life is enough”, or hearing fans, or using my laptop as a hand-warmer in summer. And I don’t understand how 1080p is still so common.
I can’t wait for M2 (pro) support for my MacBook Pro. I was long term Thinkpad/Arch Linux user and really want to go back to such setup. Sadly I didn’t find anything better hardware-wise than the MacBook but I love Linux.
I know they are focused on getting it to a good quality on M1 first but eagerly consuming all project updates! Good job team!
When the Asahi alpha released dropped last March, it was stable and functional enough to become my daily driver. And since then the pace of development has been steady and impressive - I actually look forward to every 'pacman -Syu'.
If someone told me 20 years ago that I'd be using an Apple-made ARM workstation running Linux to do all of my development work today, I would have never believed them.
It is GPU accelerated and very smooth since late last year. Battery life is about 6-8 hours on an M1 air for me, and should improve significantly with this latest upgrade (haven't measured yet).
The installation experience is very smooth, so it's very easy to try it out. You can nuke it afterwards if it doesn't work for you.
> "This release features work-in-progress OpenGL 2.1 and OpenGL ES 2.0 support for all current Apple M-series systems. That’s enough for hardware acceleration with desktop environments, like GNOME and KDE."
This gets mentioned basically every time. The gist is that:
- OpenGL is a much easier target to support, in terms of having a functional desktop. Remember, hardware-accelerated apps have been runnable on Asahi for months now, while Vulkan support is still a while away.
- Most of the work being done is common to OpenGL and Vulkan, so it's not exactly a ton of wasted effort.
Zink requires a fairly advanced Vulkan driver, with many complex optional extensions. A basic Vulkan driver wouldn't cut it. So getting basic OpenGL working is much less work, especially since they can share a lot of work with Mesa's existing OpenGL drivers.
Recently, I've been thinking about using Asahi as the host system running on my M1 MBA and run everything macOS in a vm. Does anyone have experience with that? How stupid would that be?
People have gotten QEMU with KVM running VMs and other people have gotten macOS ARM64 running under QEMU... so I think technically possible, though I haven't actually tried going that far with it. I don't think you'd get virtual GPU acceleration support working in the current state, so performance would be pretty god awful.
Does anyone know how efforts are going to to get Asahi's kernel changes merged back upstream? Has anything been merged upstream already? Is there a roadmap for getting things upstreamed? Are they pursuing an incremental approach to upstreaming individual components or do they have to prove that everything is flawless on Apple silicon before anything can be merged?
The feature table here https://github.com/AsahiLinux/docs/wiki/Feature-Support also tells you where along the path something is in terms of being stable and making it into the kernel. If it just has a release number like "5.19" then that means it was successfully upstreamed as part of that kernel version. Generally, the Asahi team tries to upstream as they go, but if they are uncertain about the overall architecture of a driver they'll hold off to prevent ossifying certain design decisions too early. An example of this is the GPU driver.
Proton is a wrapper/tweaks to wine, right? Thus x86-64 windows games can run on x86-64 linux. Thus the steam deck and many happy linux gamers. Apparently it's good enough to run a large majority of steam games, I believe I remember something like 90+% of the top 50 games on steam.
I don't think it helps at all with running x86-64 code on arm.
Wow, excellent. I'm trying to justify replacing a 2015 desktop with either a mac studio or a ryzen/zen4 desktop. I'm not a gamer, but do occasionally fire up steam to play something old like Xcom 2 or Majesty.
Knowing that at least some x86-64 steam games could work is promising, thanks.
For me it all comes down to the extra memory bandwidth.
A recent metal port of llama is pretty tempting and being able to run GPU accelerated LLMs with greater than 16GB (mid range GPUs) or 24GB (highend/RTX 4090) on the mac studio is interesting. $3,600 (for 96GB ram) interesting, not so sure.
Where many steam games where tried live and debugged to run FEX + wine + Asahi GPU driver through its paces. Highlights include Portal working, Fallout NV working. Crysis running with graphical artifacts and lots of other games. The main limitation currently is simply GPU driver feature support. To run modern D3D12 games you basically need Vulkan which Asahi GPU driver doesn't support yet.
Yes, that's right; I generally pop back in to macOS every so often to check for OS updates (which update the firmware as well - no way to just update the firmware).
There is a recovery partition (in fact, it's used when installing Asahi) but I don't think it's possible to update the firmware from there. Boot CD might work. You could try asking on r/AsahiLinux.
Awesome work. Now if we could just get those parts in a modern, modular laptop with replaceable components. Frame.work has raised the table-stakes for laptops.
I typically work on my work issued linux workstation (pretty old now and less powerful than my Macbook Pro), but often need to do development from my laptop too. I run a Fedora arm VM in UTM full screen. It generally works well, although I'd much prefer to have native linux on my laptop. It would be nice of mac os didn't interrupt full screen mode randomly and allowed UTM to capture all keys and touchpad gestures.
I personally am incredibly unproductive on a mac, and have no idea how anyone does anything with how terrible the window management and virtual desktops are. Plus, I find all my linux based tools that I am accustomed to just work so much better under linux. So for me, even a VM is still a huge leap in productivity.
I have done so for years on my 2013 x86 MBP, using VMware Fusion.
It's really nice being able to four-finger swipe between Macos and Linux full-screen desktops when I want that.
But usually I SSH into the Linux VM so that I can use iTerm2 as my terminal, and use Mac native GUI Emacs (railwaycat & Mitsuharu Yamamoto's version) to have its graphical features working alongside my iTerm2 and Firefox.
Linux in a VM runs some things much faster than the host MacOS on the same laptop! In particular filesystem intensive builds. I found some Docker-based test system I had to use for work ran an order of magniitude faster in the Linux VM than with Docker on the Mac host.
I did test running Linux natively on this MBP (no MacOS host) but found to my surprise the battery drained faster than using it in a VM, and the VM is almost as fast. So I stick with the VM and I get to use both OSes side by side.
I don't use the VMware file sharing as it is extremely buggy and regularly corrupts files when operating on them with Git or some other activities due to buggy cache settings which cannot be set to non-buggy values.
Instead I use some crazy combination of Samba and NFS which works reasonably to get the right permissions and ownership as seen by each side of the other's files. The upshot is two-way file sharing works, but this was actually difficult to make reliable.
(I found the Mac SMB2 client is prone to occasionally deleting files that aren't even being accessed by it at all, but are in a directory where other files are being updated. A shocking bad SMB2 protocol bug in the Mac client where it sends the wrong filename for renames occasionally, using a name from a recent directory listing. It took me a year to diagnose that and I thought it must be a bug in my code in a critical Linux database application I was writing, which I didn't dare deploy upstream for all that time, because important random files like invoices and data files would occasionally disappear and be noticed days later causing me to doubt myself. That is until I did a very deep dive with every debugging tool you can imagine, to find the true cause was MacOS when I edited unrelated files, and my application had been fine all that time to deploy to prod. Solution was force MacOS to use CIFS instead of SMB2.)
I hope to replicate something like that when I get an ARM Mac at some point, though I hope the MacOS SMB bugs are fixed. Currently holding out for an M3.
I'm using a multipass docker setup. It is faster than my previous setup, but when running make -j from inside the container top says that only 1 of 8 CPUs is getting scheduled to userspace at a time.
Also, the bind mount of the external MacOS directory is extremely slow. I do out of tree builds so that the builds land in ext4.
I haven't gone the next logical step because (1) it is plenty fast for rust development, and battery life is fine, and (2) I'd like webcam and speaker support.
Also, I just checked, and I have a 2023 model, and the installer status is "WIP", so I guess I'll be waiting a bit longer.
They’ve never done an M2 Ultra, as no one (anywhere) has one yet.
I believe they said they will work on it but they can’t promise a support date because they don’t know how long it will take once they have the hardware. And I don’t know when they’ll get the hardware.
So basically the recommendation was “don’t buy it to run Asahi yet, you’ll probably be waiting a bit”.
It wasn't until this morning installing Asahi on an M2 Pro was supported in expert only install. Prior to that it hasn't been supported at all. There is still some work to go on the Pro/Max/Ultra SoC family.
From a follow up post on Mastadon https://social.treehouse.systems/@AsahiLinux/110497512340479...:
"Also in this update:
We now have a cpuidle driver, which significantly lowers idle power consumption by enabling deep CPU sleep. You should also get better battery runtime both idle and during sleep, especially on M1 Pro/Max machines.
Thanks to the cpuidle driver, s2idle now works properly, which should fix timekeeping issues causing journald to crash.
Also thanks to the cpuidle driver, CPU boost states are now enabled for single- and low-threaded workloads, noticeably increasing single-core performance.
Thermal throttling is now enabled, which should keep thermals in check on fanless (Air) models. There was never a risk of overheating (as there are hard cutoffs), but the behavior should now more closely match how macOS works, and avoid things getting too toasty on your lap.
Random touchpad instability woes should now finally be gone, thanks to bugfixes in both the M1 (SPI) and M2 (MTP) touchpad drivers.
A bugfix to the audio subsystem that fixes stability issues with the headphone jack codec.
New firmware-based battery charge control, which offers fixed a 75%/80% threshold setting. To use this, you need to update your system firmware to at least version 13.0, which you can do by simply updating your macOS partition to at least that version or newer. This new charge control method also works in sleep mode.
U-Boot now supports the Type A USB ports (and non-TB ports on the iMac), so you can use a keyboard connected to any port to control your bootloader.
And last but not least, this kernel release includes base support for the M2 Pro/Max/Ultra SoCs! We are not enabling installs on these machines yet as we still have some loose ends to tie, but you can expect to see support for this year's new hardware soon."
"New firmware-based battery charge control, which offers fixed a 75%/80% threshold setting. To use this, you need to update your system firmware to at least version 13.0, which you can do by simply updating your macOS partition to at least that version or newer. This new charge control method also works in sleep mode."
This is interesting, am I correct in thinking this a feature implemented by Apple and now supported by the Asahi team? Does that mean that macOS supports this charge control feature?
I really hope Apple brings the same charge limiting to iPhone as well.
> This is interesting, am I correct in thinking this a feature implemented by Apple
Yes, battery charge control is a hardware(/firmware) feature supported on other modern laptops as well, such as the Lenovo ThinkPads, but it's not a standard so it requires explicit driver and OS support.
OpenBSD recently added support for this as well for both of these implementations (Apple silicon and ThinkPads).
https://marc.info/?l=openbsd-cvs&m=168436150408382&w=2
https://marc.info/?l=openbsd-cvs&m=168458409622780&w=2
https://marc.info/?l=openbsd-cvs&m=168521616605492&w=2
I know certain Android/Samsung phones support this as well, not sure about iOS/macOS.
I am on macos 13.3.1 and have noticed this feature for at least a couple of months, maybe longer. It says Charging On Hold (Rarely Used On Battery)
Optimized Battery charging has been available since Mac OS 11. Works on intel MacBooks too.
https://www.macworld.com/article/235001/macos-big-surs-batte...
I can never get that to activate, even though I use my MBP 14" 100% at home...
In my experience for it to take effect you need either a predictable window of plugged in time (it will delay charging until it needs to to hit 100% by when you normally unplug) or just to use your compute almost exclusively plugged in (at which point it will start letting the battery drift down to 80%). If you regularly unplug to work on the couch or whatever you may not ever see the effect directly unless you wake up an hour early to check the charge state of your laptop.
It’s in settings under battery
I really want to try this out on my work ThinkPad since it stays plugged in effectively all the time. It requires a kernel module though, I'd need to get that approved first and I haven't got around to it.
Naïve me doesn't understand why it can't be done from userspace, but I'm sure there is a good reason...
If your laptop firmware doesn't support it, here is one trick if you have a removable battery:
Put a piece of paper over one of the batteries middle contacts. That will make the firmware think the battery is overheating. It will then refuse to charge, but will still happily discharge.
You can do that to keep your battery at 80% while still on AC power. Handy if you operate from AC power 99% of the time, yet don't want your battery to die from being stored at 100% charge and hot for many years.
> Put a piece of paper over one of the batteries middle contacts.
Bad, very bad idea if you don't know what you are doing - depending on where the "smarts" in the BMS are, you may damage your battery or make your BMS think the pack is broken or prevent your BMS from recognizing a charge state mismatch (and in the worst case, a cell going undervoltage or reverse polarity) as you have a good chance that you cut off one of the cell balancer contacts. This trick only works with removable phone batteries.
Would an engineer really make a consumer design where a sliding contact failing open led to a safety incident?
Every laptop since the 90's has the balancing and protection circuitry inside the battery, not inside the laptop.
Indeed, at worst you'll get a "no battery installed" and the laptop might not even try to discharge it.
The comment you're replying to is just paranoic FUD. Which removable laptop battery has external balancing contacts!?!? They're all +/-/data.
I'm a corporate firefighter assistant. If there's one thing people should not be messing around with, it's Li-Ion batteries. These things are dangerous enough even without people messing around with their control and management systems.
> I really hope Apple brings the same charge limiting to iPhone as well.
This was added to iPhones in 2019.
> If your iPhone stops charging at 80%, it's most likely due to a feature Apple introduced in iOS 13 called Optimized Battery Charging. It aims to prevent over-stressing the battery and hence extend the battery life of your iPhone by limiting the charge to 80%.
Your iPhone learns your usage patterns and delays 100% charging until moments before you wake up in the morning.
https://www.makeuseof.com/why-your-iphone-stops-charging-at-...
It's a bit different, though. I don't often need my iphone to last multiple days on end. Yet, if I keep it plugged in as often as I'm sitting at a desk, it'll never go below 80%. If I get it below 80%, sooner or later, it will figure "i want to use it" and will charge it all the way to 100%. The lowest my battery ever got on this phone was 40 something when I was away for a weekend without a charger. It's very rare I use it a lot, so the "intelligence" clearly doesn't care how long the battery needs to last.
The way it's implemented on my mom's android, it always shows 80 or 85% (can't recall which one it is), even if she leaves it plugged in for the whole weekend.
On my HP laptop, if I activate the "battery saver mode" (as opposed to "AI"), it reports the maximum capacity as somewhere around 80% of the design capacity. I don't know whether Linux cooperates with this, but probably not. HP only talks about OS compatibility for the "AI" mode, which not only requires Windows but a specific HP app.
> It's a bit different, though.
Yes.
It learns your normal waking time (if you have one) and gives you a full charge before you wake up.
Which is what I want. A full charge for the working day, without needlessly shortening the batteries functional lifetime.
Sure, but if I know I won't drain my battery more than 10% that day, I can't tell it not to top it up more than 80%. That's the case for me 99.9% of days.
This also seems to work only if you've drained the battery below 80%. If it says 90% and I plug it in? It'll charge it fully right away.
That's not how Apple's implementation works.
It doesn't charge the battery over 80% at all, unless it detects that you have a normal waking time, in which case it charges fully right before you normally wake up.
Unfortunately, other companies copied the 80% charge bit without copying the part about figuring out if you have a normal waking time and giving you a full charge right before that.
For instance, Samsung's S23:
>Once you turn off the battery protection function, you'll be able to charge your battery up to 100%
https://www.samsung.com/ae/support/mobile-devices/battery-pr...
> It doesn't charge the battery over 80% at all, unless it detects that you have a normal waking time, in which case it charges fully right before you normally wake up.
My iPhone disagrees with you. I go to bed and wake up at just about the same time. If I somehow managed to drain the battery below 80%, when I plug it in, it'll charge to 80 and then tell me that by the time I wake up in the morning, it'll be fully charged. Which is the case.
But the most common use case, for me, is waking up, leaving the phone plugged in (WFH). Maybe go for a walk around noon, snap a picture or two. Forget to plug it back in and get back to work. Plug it in as I go to bed at 95%. It tops right up, doesn't wait until the morning.
> Unfortunately, other companies copied the 80% charge bit without copying the part about figuring out if you have a normal waking time and giving you a full charge right before that.
But that's actually what I want. With very, very rare exceptions, I never need a full charge on my phone. Hell, until a few weeks back, I was rocking an iphone 7 with a battery in a questionable state (started bulging). That thing never went below 60% with my use patterns. 60% meant night out, so a hefty dose of maps use. Normal days, it didn't go below 85-90%.
I think the issue is that Apple expects that when you wake up in the morning, you unplug your phone for the day and actually use it a significant amount. Which isn't my case at all. If I don't leave my house, it'll stay plugged in. I usually forget about my phone since everything I need to do, I do on my computer. For the occasional phone call, I can either keep it plugged in (the cord is long enough) or it's already connected to my headset, so don't even need to go fetch the phone. If I don't go to work, it can stay plugged in for days on end (I don't always grab it when I go out).
> that's actually what I want
Why protect your battery's ability to hold a full charge over a longer lifespan if your battery is constantly throttled to a partial charge anyway?
Because it's my understanding that charging and keeping it all the way full shortens significantly its battery life compared to 80-85%. I would rather not have to dick around plugging and unplugging the phone to not have it be full all the time. Plus, that also counts as consumed cycles.
I've used my old phone for a good six years. I've swapped its battery some three years ago, and it would've needed a new one now, had I continued using it.
If that phone is any indication, my current one should be in service for at least as long. If I can avoid having to swap its battery, it's a win in my book.
It can also sometimes happen that I foresee being away from an outlet or otherwise need as much charge as possible. In those situations, I'd just deactivate the battery saver feature and let it charge to 100%. So if the remaining capacity is closer to its original one, again, it's a win.
> Because it's my understanding that charging and keeping it all the way full shortens significantly its battery life compared to 80-85%.
If you don't charge fully until just before the user's normal wakeup time you aren't keeping it all the way full all the time.
You protect the battery lifespan and get a full charge at the beginning of the day, instead of having a battery that constantly holds less charge.
> If you don't charge fully until just before the user's normal wakeup time you aren't keeping it all the way full all the time.
You keep repeating this and seem to ignore my observation that if the battery isn't drained below 80% when I plug it in, it will recharge it fully immediately. It will not wait until the user's wake-up time.
And in my case, it's rare that the battery falls below 80%, so whenever I plug it in, it gets recharged fully right away.
So, in practice, it's all the way full all the time.
> in my case, it's rare that the battery falls below 80%
Yet year after year, surveys show that more battery life is the feature people want in a phone the most.
> 9 out of 10 phone users have low-battery anxiety
https://electrontogo.com/blog/9-out-of-10-phone-users-have-l...
Intentionally throttling the battery's ability to charge no matter what is going the wrong way.
My Sony Xperia 10 IV lets you set it to never charge above 90 or 80 %, as an alternative to it learning your habits. I have it set to 80%, and I've been unable to use it up in under 2 days. I've heard that iPhones have similarly good battery life as the 10 IV, so it seems, to me, that get quite far on 80%.
I bought the same phone recently, and for my light usage pattern (some texting, the browser, Google Translate, Google Maps, a calculator app, and the camera), the battery lasts almost 5 days!
I had completely forgotten how it is to own a phone that doesn't need to charge everyday...
> I've been unable to use it up in under 2 days.
Take up playing a resource intensive game like Genshin Impact and you can very easily drain the battery in a day.
> 60fps highest settings 100% brightness Low sound Home WiFi 3:20 total
https://www.reddit.com/r/SonyXperia/comments/onb3dw/genshin_...
> This is interesting, am I correct in thinking this a feature implemented by Apple and now supported by the Asahi team? Does that mean that macOS supports this charge control feature?
It does, but in a weird way. You can turn on "adaptive charging" and it will randomly decide to charge to 80%.
If you want to properly control it, just install the wonderful AlDente utility ( https://apphousekitchen.com/ ). Then you can manually control the max charge percentage. Mine is permanently set to 80% because I never really use even 40% of the battery on my M2-based laptop.
I also appreciate the detail in which the Asahi team presents the progress they have made.
I do not follow Apple's release notes so I cannot compare.
The Apple release notes for macos look like this https://www.apple.com/macos/sonoma-preview/
Such an amazing set of engineers jacking away at this. What awesome work they’re doing.
Huge respect for these people. Wish I could do what they do.
ugh ruined by autocorrect
s/jacking/cracking
I thought you meant like a jack hammer.
Yup. That’s exactly what I meant.
Some engineers are probably of that persuasion too. I'm not going to judge. LOL.
smirking as an engineer myself I can only confirm as much haha.
>And last but not least, this kernel release includes base support for the M2 Pro/Max/Ultra SoCs! We are not enabling installs on these machines yet as we still have some loose ends to tie, but you can expect to see support for this year's new hardware soon."
Amazing
> e includes base support for the M2 Pro/Max/Ultra SoC
Does this mean Apple gave them prerelease hardware early? Might apple start helping these guys more - like for example donating a 5 person dev team for a few months maybe?
It means that Apple isn't radically changing the internals of the SOC every year.
>Apple’s first iPhones ran on Samsung SoCs, and even as Apple famously announced that they were switching to their own designs, the underlying reality is that there was a slower transition away from Samsung over multiple chip generations. “Apple Silicon” chips, like any other SoC, contain IP cores licensed from many other companies; for example, the USB controller in the M1 is by Synopsys, and the same exact hardware is also in chips by Rockchip, TI, and NXP. Even as Apple switched their manufacturing from Samsung to TSMC, some Samsung-isms stayed in their chips… and the UART design remains to this day.
https://asahilinux.org/2021/03/progress-report-january-febru...
M2 Pro/Max were available in January. I think they needed to wait until now to be sure the M2 Ultra announcement didn't have too huge of changes from the way the M1 Ultra was done. In other words, the Asahi Linux team don't have an M2 Ultra to test on, they are getting ready for when they can get some test results, possibly from users.
Please consider donating if you have the means. https://asahilinux.org/support/
Another follow up on Mastodon is about the ban evasion that HN does specifically for Asahi. https://social.treehouse.systems/@marcan/110503050993279759
I find it pretty tasteless for HN to do that.
It’s a little hard to understand what is going on. Can you explain? Thanks.
Asahi Linux's site previously redirected any site with the Referer header set to 'news.ycombinator.com' to google.com. They did this because they believe that Hacker News' community fosters a hostile environment, and thus it was blocked. Setting `rel=noreferrer` on an `a` tag means that the browser will not submit this header, thus meaning this sort of redirect will not occur. Also(as per Marcan's Mastodon account), this rel tag has only been added to asahilinux.org links, not any other links on HN. This means that the moderators of HN believe that the commentary that has in the past been quite misinformative should be allowed, and they are deciding to bypass the block that was put in place by the Asahi team. Now, as per https://social.treehouse.systems/@marcan/110503331622393719 , they're adding an HTML header to anyone who has submitted a post to HN(or has the submit page in their history), as they have no other way of doing anything, as they no longer have the Referer header.
Doesn't seem to be applied to older posts https://news.ycombinator.com/item?id=33891197
This raises interesting moral questions that I'm not sure I have an answer to.
For one, I don't believe this place fosters a hostile environment, although it's definitely a place where people love to tell you how you're technically wrong about something.
For another, I would guess that there are a very limited number of websites that would opt into this sort of anti-traffic behavior. Hacker news could certainly choose to honor it, but it also feels within their right to bypass the block.
I wonder if there's a sort of middle ground, where HN alerts the user of the redirect that would have occurred, but still shoots the user to the desired location.
But now you're getting into user flows and begging the question as to why the redirect is there in the first place.
I'd love to know more about the perceived hostility, even reading up on Mastodon left me with more questions than answers.
there is constant transphobia against Asahi contributors on display on this site. Every single article that gets to the frontpage will have dozens of vile posts in the deads and a couple sprinkled inbetween. I’m not sure if HN staff doesn’t care about it, agrees with them or simply isn’t aware.
It’s not limited to the Asahi project, either. The site isn’t safe for queer folks.
I had a look at the comments on this post. I don't see any saying anything at all about gender, pronouns, trans-ness, etc. (Other than in this little discussion right here about what is allegedly wrong with HN.) I have showdead on, incidentally; I'm not seeing any dead comments on those topics either.
Maybe for whatever reason this particular post is better than others? I put <<asahi site:news.ycombinator.com>> into DuckDuckGo. (Not because there's anything magic about DuckDuckGo search results, but to get a sample of Asahi-related HN posts without cherrypicking.) The first hit is to https://news.ycombinator.com/item?id=35394297; no trans-related things there that I can see either. (It does have some discussion of why the Asahi Linux site was turning away people coming from HN, and the reason given there was nothing to do with "Lina", nor with transphobia, but was that marcan feels that HN discussions of Asahi are full of mistakes.)
Second hit is to https://news.ycombinator.com/item?id=35928000. Again, nothing trans-related.
(To look for trans-related things I (1) skim-read the comments by eye and then (2) searched for "trans", "gender", "pronoun" and "man", the last because people being obnoxious about trans issues often can't resist going out of their way to call someone a "man" or a "woman" if they think they might be upset by being called that. I specifically kept an eye out for dead comments. It's very possible that I missed things, but it seems to me that a minimum threshold for saying that there's "constant transphobia on display" is that someone explicitly and somewhat carefully looking for it should be able to find at least one example.)
Next few links aren't obviously strongly-Asahi-related articles. Next one that is is to a comment about "Lina" somewhere in the middle of https://news.ycombinator.com/item?id=35233479, so I took a look at the comments there. There was indeed a "Lina" subthread, some of which was pretty rude but none of it in a visibly-trans-related way.
My general impression of the HN comments on Asahi-related threads, from this, is that they are on the whole very positive, that some people think marcan is weird for Lina-related reasons (which may or may not actually make sense; I have no knowledge of that business), and that if anyone is being obnoxious at or about trans people in those threads then either they're doing it in ways I'm failing to see or else it's being cleaned up effectively enough that it's gone by the time I look.
It seems your experience is very different. Where should I be looking for some examples?
> Every single article that gets to the frontpage will have dozens of vile posts in the deads and a couple sprinkled inbetween.
But that's why they're dead. Turn off "showdead" if you don't want to see them.
The ones that aren't dead, I would assume the moderators haven't seen yet and/or haven't received enough flags yet. Moderation is hard.
I mean, every time Foone asked to be respected and not posted here HN staff didn't seem to care...
I’m confused by this comment. I follow the asahi Linux developments here on hn and read the comments. In this thread alone there are several high rated posts to the effect of “this project rocks!!”
The only reference I can find to trans- anything is basically this comment. Quite honestly, I recommend a thicker skin. Some people will be assholes, whether you’re trans, straight, bi, whatever. Pretending that assholes don’t exist will just make you angry… you can’t just wish away the assholes but you can ignore them and prove them wrong by being awesome (which in this case the asahi linux developers work speaks for itself!)
It seems HN's philosophy is like that of (early) Reddit, where the commenters self-govern and down/flag comments like that - and it seems to be working if they're all dead, right? They're hidden unless you have an account and specifically enable showdead.
> vile posts in the deads and a couple sprinkled inbetween. I’m not sure if HN staff doesn’t care about it, agrees with them or simply isn’t aware.
So basically you're judging a community by the posts the community and moderators deemed inappropriate and moderated away? Rather curious metric.
If you don't want to see dead posts then turn showdead off (which is the default). That you get some transparency in the moderation is a feature, IMHO, but it's also opt-in.
Maybe there should be a "verydead" for these kind of outright idiotic posts so they just won't display at all for anyone, but then how do you prevent abuse and keep the mod workload reasonable marking endless posts as "verydead"?
[flagged]
> there is constant transphobia against Asahi contributors on display on this site
Could you be so kind as to link me to an example?
People have discussed Marcan's anime alter ego, certainly, but I've never seen anything transphobic in the slightest. Maybe it's all flagged before I get to see it, but I honestly can't figure out what you might even be referring to. I enjoy reading Asahi updates when they come up, so I am unsure how I could be missing something so 'constant'.
> It’s not limited to the Asahi project, either. The site isn’t safe for queer folks.
As a 'queer person' myself, I find that statement utterly ludicrous and I reject it completely.
> People have discussed Marcan's anime alter ego
This is pure and poor conjecture, just like the rumor (originating on chan boards and KF) that a person close to marcan had faked suicide. Exactly this sort of rumor mill is what is wrong with HN linking to Asahi developers.
> As a 'queer person' myself, I find that statement utterly ludicrous and I reject it completely.
Are you a cis gay? I remember a few cases where people reached out to have information redacted to dang and it took weeks. While the people mentioned above were digging into private lives. This site absolutely requires you to shut the fuck up about your own life if you are at risk of being turned into a "lolcow" or pedojacked or whatever else these people will come up with. I understand that's not a consideration yet for 'queer people' currently not conscripted to the front of the culture war.
I can honestly say I've never, ever, heard anyone call anyone else these terms on Hacker News. I don't think I've ever even heard those terms before. It sounds like you might be very online, in a way that seems to be making you unhappy.
Some of my most upvoted comments on here have been from a 'queer' perspective. It helps to assume good intentions and engage with others constructively and in good faith.
I must admit, I feel saddened by your dismissiveness of certain 'queer people'. You have no idea who I am, what my identity is, and yet you so casually dismiss members of the marginalised group you purport to be defending. How callous of you.
Perhaps the reason I can so easily dismiss your hysterical claim that HN is unsafe is that - in my day - 'unsafe for queer people' meant 'reasonable likelihood of getting a brick to the face', and not 'seeing words online you don't agree with'.
this is very dismissive for how much you're projecting i'm dismissing you. it's okay if you can't relate to me, just don't assume you understand.
hn is unsafe in the way that it provides zero control over what you have submitted and a perfect history of your posts, mostly set in stone unless you email an administrator. at least reddit lets you self-service delete posts.
those are perfect conditions for a certain kind of group that spend way too much time digging up and correlating info to then start harassment campaigns that exceed 'seeing words online you don't agree with' quite often.
you can of course now go on to scold me and others with this problem about how we need to up our opsec or shouldn't post in the first place. i find such arguments, if you were to make them, entirely unconvincing. being aware of your risk profile is one thing, shifting all the blame for making it harder to retroactively rectify little pieces of information (these people found a place from a blurry 500x500 picture of a parking lot out a window) on the user is just a bad excuse for shitty UX.
> you can of course now go on to scold me and others with this problem about how we need to up our opsec or shouldn't post in the first place
Not at all. I neither think you should up your opsec nor avoid posting. I think you should be unabashed of what you have to say. I also think you should hear out others doing the same, generously and in good faith.
That tiny minority who harass and abuse queer people today? That used to be virtually everybody, all the time, everywhere you'd go.
The only reason queer rights are where they are today is because people weren't afraid to speak up, even when they had every right to be. When coming out meant admitting to criminal acts - proclaiming them in public, no less. Had they not put themselves out there - had they refused to speak to those that didn't already agree with them - this would be a much darker world, and your definition of 'unsafe' would be a lot more visceral.
People used to march, faces out in the open, in their small towns, for their rights, past neighbours who hated them. That's what 'unsafe for queer people' means. I'm just never going to be able to see 'people disagree with me online sometimes' as being in any way comparable.
A key tenet for many people appears to be that they should be able to do whatever is within their rights, but that leaves out the fact that sometimes doing something within your rights makes you an asshole.
I see the HN admins failing this moral test.
> This raises interesting moral questions that I'm not sure I have an answer to.
It's not really all that much of a moral conundrum. Marcan's belief - expressed a number of times on his Mastodon - appears to be that he can prevent other people from discussing something, for the sole reason he doesn't want it discussed. It's not a particularly defensible position in an open society.
In particular, he is upset that people on Hacker News tend to point out that a contributor on Asahi - Lina - appears to be a computer-generated anime alter ego of Marcan himself.
Me, I have absolutely no problem with Marcan having an anime alter ego, but I don't think it's entirely reasonable to expect people to refrain from noticing this and remarking on it. Marcan disagrees, and this is the source of the HN-Marcan rift.
(As I've remarked before, I do mind OSS projects listing fake contributors, for both ethical and legal reasons, but that's another discussion.)
Quit with the gaslighting. It's not about an anime avatar, and you know it.
The problem is more that HN is perceived, with good reason, to be transphobic, and Asahi has several trans developers. This is also the reason why a couple of other people I won't mention don't want HN to link to their projects either, because for every link, several people will bitch in the comments about the validity of their gender and pronouns. (As if creating a web app somehow makes one an authority in biology.)
I mean, obviously, it's not like everybody who comments here takes part in the abuse, but when you spend every damn day of your life seeing it and depending on where you live, possibly being harassed by the government over it too, HN turns out to be one more place where you simply can't have any peace.
(And honestly, even if Lina is a anime alter ego, which you haven't proven, and only suspect, why the fuck do you care? Don't you have something better to spend your time on?)
> gaslighting
> and you know it
I'm always so knowledgeable in the eyes of people who criticise me!
No, I didn't know Asahi had a single trans developer. I've never seen it come up on HN, where I most often hear about the cool work being done by the Asahi team. The vast majority of comments on here about every new Asahi article are effusive praise.
I dispute the characterisation of HN as some transphobic hellsite. That characterisation is simply not accurate, in the slightest, whatever perception may prevail on the Asahi Discord.
> why the fuck do you care
My other comment below, explaining the legal issue with fake OSS contributors, was already up by the time of your comment, so I refer you there. More broadly, I think Marcan should get to have as many anime alter-egos as he wishes. I'm just not particularly surprised that people find that noteworthy, and I think if you choose to have a public anime alter ego, you probably should be able to deal with that? I feel like Marcan's attempt to shut down this discussion is a perfect example of the Streisand effect - I certainly would have never found out about 'Lina' were it not for this silly feud.
> several people will bitch in the comments about the validity of their gender and pronouns.
I've seen these posts, and they're horrible, mean-spirited, and hurtful.
But as far as I've seen all those posts have also been downvoted to hell, flagged, and hidden, and the users often banned, and they're often from new "green" users rather than regulars. It's entirely possible that some of these posts remained (especially if they're posted after the conversation died down and there are less eyes), but I'm certain that if you email dang that he will take action.
At some point the first comment on one of my articles was "gay n---r soiboy" or something to that effect from a new account – a curious comment since my website has a picture of me being white enough to get a sunburn in Ireland – but how do you prevent that? Limiting sign-ups is the only thing I can think of.
I can definitely understand that people feel very negative about these things, but I think it's unfair to judge all of HN by it – you're essentially judging a community by the posts that were considered inappropriate and were removed.
And if you don't like HN (for any reason) then that's fine, so don't visit HN then. All this tomfoolery with referer blocks seems rather at odds with how the internet is supposed to work; "microsoft.com blocks links of the referer is from lwn.net" would cause a loud uproar here.
A good comment on this at last.
[flagged]
You are misunderstanding. You can discuss your conspiracies anywhere you want, just not on a post pointing to the people that have to deal with the fallout of your unfounded conjecture.
Also assuming there is a "fake contributor", who cares under which names contributions are split up? The work still got one. Also, it is absolutely not your project, you don't get to demand people show their ID when they write code for a project.
> Also assuming there is a "fake contributor", who cares under which names contributions are split up? The work still got one. Also, it is absolutely not your project, you don't get to demand people show their ID when they write code for a project.
If I contribute my code to an open source project, then I - as the copyright holder - agree to license my work under an open licence.
If I use an OSS project, I am using other people's copyrighted work under an open licence from them. Without that licence, I have no legal basis on which to use that work.
Only real people can (currently) hold copyright. If person X writes some code, but the licence (incorrectly) attributes the copyright to person Y, and person Y purports to give me an open licence to use that work, then - crucially - I have no license from the actual copyright holder (person X) to use their work.
Until an effective open source license is made, this code is not open source; it is completely proprietary. If person X chooses to sue you for copyright infringement, it is no defence to say that you're using it under a license from person Y, because person Y had no right to give you that licence.
This is a major ethical and legal problem. I would be very wary of the Asahi codebase.
If you wanted to split hairs this thin, you wouldn't use any project with at least one german citizen as contributor, since they can never truly yield all copyright on a work. You'll be fine. Anonymous contributions to free software (or even entire releases done anonymously, e.g. Bitcoin) are not actually uncommon.
As I said, very weird hangup to have, definitely not motivated by other reasons.
> If you wanted to split hairs this thin, you wouldn't use any project with at least one german citizen as contributor, since they can never truly yield all copyright on a work.
Nor can anyone else, copyright is not generally 'destructible'. That's why it's a licence. The holder keeps the copyright, but licenses the work to the general public. (Assignments are another way to achieve something like this, provided the assignee then licenses the work.)
I assume what you're referring to is inalienable moral rights - hence the reference to Germany - but those are a feature of many (most?) of the world's legal systems. They are included in one of the revisions to the Berne Convention, if I recall correctly, which is an international treaty on intellectual property.
I understand you're sceptical, but the legal dimension of OSS does matter. Using copyrighted material without a licence would constitute a major business risk. I would appreciate it if you could kindly refrain from making ungenerous assumptions about the intentions of others.
What makes ze Germans special?
Right to moral objection.
That doesn't explain anything. Can you elaborate?
The most commonly cited example is a musician objecting to use of their music during a neo nazi rally. They won that case as the court judged the integrity of the work to be compromised.
I see. Well, I don't think that would affect code in the same way. If they have already contributed it then it's out in the open, and unless the maintainer was doing something objectionable they wouldn't have a case, and even if they did it would only be enforceable in Germany.
> that he can prevent other people from discussing something
And again, when some rando says "they want to prevent people from discussing, free speech!" the topic is always the same, they want to be freely racist, homophobic, whatever. Cmon man, just stop.
Forced participation has no place in open society. In such a society, when in private spaces, the valid response when told "You are not welcome here" is to leave and not to harangue the host about open societies.
Adding noreferer tag is “tasteless”? What’s tasteless is jwz for example silently redirecting visits from this site to goatse. This just reminds me of the clueless businesses in the early days of the web asking for “license agreements” to link to their content.
Its been more than a year I'm running asahi on my macbook air and I can't stress how grateful I feel for enjoying such wonderful freedom.
I don't feel like ever going back to x86 to be honest, at this point there is nothing lacking or unable to run and when the neural engine drivers come online now that the GPU is starting to mature people will be able to juice out every last bit of computation this machine is capable of.
For the record, I've switched to the edge branch a couple of months ago and honestly I noticed no actual difference in my day-to-day tasks which is really telling about how powerful even the M1 is when it can handle software rendering in such an effortless manner coupled with anything else running.
Really thank god for asahi being a thing.
> at this point there is nothing lacking or unable to run
Sure there is. You just haven't run into it yourself.
Faster, cooler and more power efficient hardware is great. I just don't think that it makes up for depending on a small team of volunteers to resolve all hardware issues in an ecosystem hostile to OSS, which might break at any point Apple decides to do so.
And the incompatibilities with ARM are not negligible. If all your software runs on it, great. If not, good luck depending on yet another translation layer.
I'm sticking with my slow, hot and power-hungry x86 machines with worse build quality for the foreseeable future. The new AMD mobile chips are certainly in the ballpark of what Apple silicon can do, so I won't be missing much.
> depending on a small team of volunteers to resolve all hardware issues in an ecosystem hostile to OSS, which might break at any point Apple decides to do so
You are describing how most OSS software has been developed. I don't see how this is any different than early linux when no hardware manufacturers had any interest in supporting it.
A lot of the work that the asahi team is doing is just fixing Arm issues in the linux kernel (and sadly user space). That work will benefit everyone using Arm systems, not just folks running asahi on Apple hardware.
Its good for there to be more hardware architecture competition! I'm glad I can run my server workloads on the Arm servers in AWS that are 20% cheaper than the equivalent x86 machines. I'm glad that I can run the software I like (linux) on legitimately nice hardware (m2 air). You can make different decisions on what architectures are best suited for your needs, but the competition in the market improves the options and prices for everyone.
I've been using Asahi since the fall of 2022. When I first started using it a lot of software was broken because of bugs in that software that had never been exposed before (specifically around page sizes larger than 4k). All of that software has now been fixed. Support for linux/arm will only continue to improve as more people use it.
> I don't see how this is any different than early linux when no hardware manufacturers had any interest in supporting it.
It's very different. Hardware manufacturers in the 90s were incentivized to support Linux to expand their customer base. A trillion-dollar corporation has no incentive to sell their hardware to a niche of a niche of technical users who are not part of their software ecosystem.
Another major difference is that Asahi is a small team of dedicated volunteers who want to run Linux on their Macs. They're a niche intersection of Linux hackers and Apple fans. Unsupported hardware in the 90s typically had a much larger customer base and group of hackers willing to spend time adding supporting for it.
Even worse: Apple can decide at any point to make their hardware much more difficult to support. Newer models or firmware updates might break things. Being at the whims of a corporation that is the antithesis of F/LOSS to run Linux on their hardware doesn't inspire confidence.
> Its good for there to be more hardware architecture competition!
> Support for linux/arm will only continue to improve as more people use it.
Agreed. I'm glad that Asahi exists. But we've had ARM Linux distros for decades now. What Asahi is doing is specifically to support Apple hardware. Some improvements will trickle out to improve general ARM support, but this points out the gargantuan task they're actually up against. Not only do they need to reverse engineer the hardware, they have to resolve all software issues with Linux and ARM. My hat's off to them. The willpower, patience and skills required to wade through the absolute mountain of issues must be astronomical. Yet this is also part of my concern; how long can a developer keep the motivation and sanity to swim against the current?
It's great that Asahi works for you and everyone else. I'm just pointing out why it will likely never be my choice for any serious work.
> Hardware manufacturers in the 90s were incentivized to support Linux to expand their customer base.
That is not an accurate description of linux support by hardware manufactures from that time period.
> Unsupported hardware in the 90s typically had a much larger customer base and group of hackers willing to spend time adding supporting for it.
I also don't think this is generally correct. Have you looked at all the random drivers in the linux kernel for niche hardware. A ton of that is from one or two hobbyists taking the time to add support.
> Apple can decide at any point to make their hardware much more difficult to support. Newer models or firmware updates might break things. Being at the whims of a corporation that is the antithesis of F/LOSS to run Linux on their hardware doesn't inspire confidence.
I guess, but so what? Apple can't break the hardware they are already shipping if you are just running linux on it. Its true, I might not buy a theoretical future laptop from Apple if I can't run linux on it, but I don't see how that would affect my purchasing decision for hardware that is currently available.
> The willpower, patience and skills required to wade through the absolute mountain of issues must be astronomical. Yet this is also part of my concern; how long can a developer keep the motivation and sanity to swim against the current?
Hmm, maybe you've not worked on projects like this, or are motivated by different things. To me, reverse engineering a thing to figure out how it works and then writing software to get it to do things the original designers hadn't planned for is one of the more satisfying and fun activities of being a software engineer. I suspect the asahi team is having fun doing a lot of this work. (That's not to say its all fun. It sounds like getting things upstreamed has been trying. I also think having to read giant comment threads where people are needlessly negative about their work might be a bit demoralizing.)
> It's great that Asahi works for you and everyone else. I'm just pointing out why it will likely never be my choice for any serious work.
You should obviously run whatever works for you.
> Have you looked at all the random drivers in the linux kernel for niche hardware. A ton of that is from one or two hobbyists taking the time to add support.
Sure, for _niche_ hardware. When was the last time a GPU driver was added by reverse engineering it? The single Nouveau maintainer was burnt out, last I heard, and the project was never a serious alternative to NVIDIA's closed driver. Kudos to whoever found the energy to contribute to it, but these projects usually don't have a bright future.
Now expand that to include maintaining all Apple devices, and it's an insane amount of effort realistically unsustainable for any group of volunteers. But good luck to the Asahi team.
> Nouveau maintainer was burnt out, last I heard, and the project was never a serious alternative to NVIDIA's closed driver
Simply not true. I recall in the mid '10s using it because the proprietary driver was crashy garbage. No, I didn't get the same performance possible with the proprietary driver, and I didn't have a bleeding-edge video card, but it was more than usable as a daily driver.
>When was the last time a GPU driver was added by reverse engineering it
Freedreno (for the Adreno family) Etnaviv (for the Vivante family) Panfrost / Bifrost (for Mali)
All these RE efforts built on each other, although the GPUs are different the tools built to do the RE were shared (and I think ashai is benefiting too).
AFAIK Google has now hired Rob Clark the Freedreno maintainer who started all this to work on Freedreno for Android / ChromeOS
Upstream Linux now has pretty good GPU support for all the major mobile GPUs these days. The hold out has been PowerVR but they are now working on an official (not reverse engineered) driver.
Nvidia/Nouveau is a good example. I've had a number of laptops with nvidia graphics. For most of that time Nouveau was _more stable_ than the official nvidia drivers. Linux clearly was a second class citizen for nvidia for most of the last 20 years. Maybe go back and rewatch linus' rant about nvidia if you need a reminder about how terrible they have been historically.
Nvidia now only sort of cares about linux because of gpgpu applications. They still clearly don't care about gaming on linux; or desktop stability.
Yes, I will take Nouveau over the official drivers whenever I can.
In the same vein I used to consider intel drivers to be the most stable and best for desktop use. Now understand there are many caveats to this statement.
I use a lot of openbsd and you won't be running nvidea drivers on openbsd for love or money. there is nouveau and they are doing amazing work, however, they are also up against a petty, secretive company that appears to hate them. So nvidea is out.
Understandably openbsd gets zero support from the manufactures. So we need an opensource driver and some brave heroic soul to volunteer their time to get it running.
AMD drivers worked well enough and if you want decent 3d acceleration the only real choice. However they tended to crash and 3d acceleration is usually not a priority if using openbsd. Also starting with amdgpu the drivers got big, really big. The amdgpu driver nearly has more code then the rest of the openbsd kernel[1]. it is this big mess of generated code where each card uses a slightly different ISA. I understand why having a stable ISA is not a priority for AMD(it lets them change the card architecture easier) however sometimes i wish it were documented and pinned down. it would certainly make for a more stable driver that is easier to integrate.
And then there is intel, Note that I have not used intel graphics since 2016 so my experience is out of date. but once you got past the first generation of intel graphics the experience was rock solid, the drivers always worked well for me. if asked for the best openbsd experience I would recommend intel every time. However my last few machines have been amd and unfortunately there is no intel graphics add-in card(i looked). My last intel box I had an amd 3d card but I only used it under windows to play games. for work/openbsd I would just use the onboard intel graphics as they were more stable.
1. https://flak.tedunangst.com/post/watc
See, now I just think you're gaslighting me.
Nouveau has never been more stable, or nearly as performant as official NVIDIA drivers. I've had the exact opposite experience from you on every laptop I've had since Nouveau was released, so we must live in different universes.
> Linux clearly was a second class citizen for nvidia
And Linux is not even on the radar for Apple. :)
Anyway, I think we've exhausted our arguments here, and are just talking past each other now. Have a good day.
It depends on what device you have.
On some of the older devices, nouveau actually works very well right now. On these devices the open driver is as stable and performant as the closed driver and has all of the same features with the exception being GPU compute stuff.
I think mostly people misunderstand the limitations of the nouveau drivers and just assume they are bad in general, but it's very much dependent on the device.
One of the big issues right now is that nouveau can't bring the newer GPUs out of "idle mode". Nvidia has explicitly restricted this feature and the chances of the issue being resolved without cooperation from Nvidia is very low. I think a lot of people try nouveau on the effected GPUs and have horrible performance and then assume it's because nouveau is bad.
I have had an interesting experience with nouveau on the GT 710. For a little while there was a bug that would cause sway to crash back to tty, at first it would happen maybe once a day, then it became so bad that it would crash as soon as sway was launched. Now in kernel 6.3.3 it seems to be working flawlessly, which is how it was at some point in the past too.
Right now nouveau is working great for me though, so it is possible for nouveau to compete with the closed drivers in specific cases, but in general, for more modern GPUs it will have very low performance due to the reclocking thing.
It's not gaslighting: https://nouveau.freedesktop.org/PowerManagement.html
Nouveau is an awesome project, but for later cards it's basically a dead project. You can get some features to work, but without proper power management there's no justifiable reason to daily-drive it. The proprietary driver is by no means perfect (particularly for Wayland) but it's the only real option if you own a modern card.
> And Linux is not even on the radar for Apple. :)
They must be awfully curious about why Xserve failed, then.
Claiming that Nouveau is or ever was as stable as the proprietary nvidia drivers feels like gaslighting to me, and I think that's what GP was saying. Nouveau has always been flaky.
If you've had good experiences with the nvidia drivers, thats great. My experiences with them have been bad and I will use Nouveau for general desktop environments unless I'm also doing gpgpu.
> the project was never a serious alternative to NVIDIA's closed driver
Because it can't properly clock the card since nvidia uses DRM to lock it out. Were it not for that, I'd be using nouveau every single day. In my experience it's more stable than the proprietary driver.
> reverse engineering a thing to figure out how it works and then writing software to get it to do things the original designers hadn't planned for is one of the more satisfying and fun activities
I feel the same way. I made a small driver for my laptop's keyboard lights and it was one of the most fun projects I've ever made. I can't even imagine how satisfying it must have been for the Asahi developers to get OpenGL running on that hardware.
This feels... fairly ahistorical.
> Hardware manufacturers in the 90s were incentivized to support Linux to expand their customer base.
Outside of some server niches (though even then, mostly not til the noughties), not really. You generally weren't looking at much first-party support at all in the 90s.
> Unsupported hardware in the 90s typically had a much larger customer base and group of hackers willing to spend time adding supporting for it.
This... is really not the case. The M1/2 Macs are _close_ to being one platform with something on the order of a hundred million units out there. Hardware in the 90s was pretty diverse; if you wanted support for your video card, say, there might be twenty chipsets out there, with maybe fifty manufacturers. Later on, there was a fair bit of consolidation, with virtually every computer, say, ending up with pretty much the same sound card chipset, but not in the 90s.
> Apple can decide at any point to make their hardware much more difficult to support. Newer models or firmware updates might break things.
Yeah, again, that _very_ much happened in the 90s.
> You generally weren't looking at much first-party support at all in the 90s.
I wasn't making a historical claim, but a philosophical one. Hardware manufacturers in the 90s had that incentive, if they wanted to pursue it. To Apple, a trillion dollar corporation, selling a negligible amount of laptops to a niche crowd of tech enthusiasts who will never become part of their software ecosystem is just not worth the effort to even pursue.
> This... is really not the case. The M1/2 Macs are _close_ to being one platform with something on the order of a hundred million units out there.
Right. And how many of those buyers are also Linux hackers willing to dedicate their time to work on projects like Asahi? How many of those would even be willing to run Linux instead of macOS? We're talking about an extremely small community of users compared even to the small Linux community in the 90s.
> Yeah, again, that _very_ much happened in the 90s.
Never said it didn't. Except that Apple is known for locking down their products, so the good faith they're showing now with leaving Macs relatively open can disappear at any moment.
> Even worse: Apple can decide at any point to make their hardware much more difficult to support. Newer models or firmware updates might break things. Being at the whims of a corporation that is the antithesis of F/LOSS to run Linux on their hardware doesn't inspire confidence.
That is the case with every single PC maker who doesn't ship all their PCs with Linux.
I just don't understand why Apple has to be treated differently. They just went through a traumatic transition where they could have locked their computers very tightly, preventing you from booting anything but macOS. They did no such thing. They did the reverse. No they didn't formally support it, but they own their own OS. How could they be expected to?
I think you're underestimating how easy it is for a corporation to change it's narrative from "we support Linux" to "we don't care about Linux".
Whatever hardware you're using, it's not like the company has signed a contract to support Linux for life.
>A lot of the work that the asahi team is doing is just fixing Arm issues in the linux kernel (and sadly user space)
While I don't have Apple hardware so haven't been closely following Asahi I dont't think that is true. Linux has supported Arm for years (more like decades) now. They've been doing excellent work on support for Apple specific hardware sure, generic Arm not so much since it was mostly done.
Let me be more specific. There were a lot of bugs specifically related to non-4k page size architectures. Arm doesn't dictate the page size so most of the Arm systems out there have defaulted to 4k pages. The Asahi wiki has a partial list of userspace programs that were (or still are) buggy and broken because they made simplifying assumptions about how different architectures work[0].
Maintainers of other non-x86 architecture have said that this is improving things for them[1].
[0]: https://github.com/AsahiLinux/docs/wiki/Broken-Software [1]: https://www.talospace.com/2022/03/asahi-linux-gives-hope-for...
If you follow Hector Martin the lead Asahi dev he's encountered a number of bugs and race conditions in ARM linux systems which were never previously exposed because there weren't blazing fast ARM chips out there that could trigger them.
Point me to the Apple contributed drivers in the kernel please.
Why does Apple need to contribute to this work to make it somehow legitimate or good? I own some nice hardware (an m2-air), I want to run Linux on it. Asahi allows me to do that! Why can we not celebrate that the asahi team is bringing oss to new hardware?
They don't need to. I keep seeing that Apple does no less than other companies w/ regard to Linux. Well- where are their kernel contributions then? Lenovo and Dell (my two laptop manufacturers) contribute.
https://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux...
> I keep seeing that Apple does no less than other companies w/ regard to Linux.
Where did I make that claim in this thread?
> in an ecosystem hostile to OSS
> You are describing how most OSS software has been developed.
Nope. Disagree here
Cool, thats a different claim than what you said above but at least one I can actually engage with.
I've run linux on many Dell and Lenovo systems over the last 25 years. Most of those systems were fully unsupported by the manufactures for anything but windows. And yet, random people on the internet contributed to make that hardware (mostly) work. I've not seen any particular improvement in the driver situation since Dell started selling linux certified systems.
Its not really surprising though, Dell is just an integrator. All they do for their systems with linux pre-installed is to pick hardware that already has drivers. They took a little bit of work out of needing to research if a given configuration is likely to work or not with linux (which is good). They don't really deserve much credit beyond that.
Its also a little funny because most drivers from hardware manufacturers suck. I don't know why, but most hardware companies are terrible at writing software. Its easy to list off hardware companies that have a long history of shipping mediocre, buggy linux drivers: nvidia, amd, broadcom, realtek, (maybe i should just list every nic and wireless chipset manufacturer). Some of these companies have gotten better and have learned how to be good kernel contributors, but they were mostly bad for years and years. Thankfully in some of those cases random people on the internet reverse engineered the hardware and contributed from scratch drivers to the kernel. Most of the time I've been happier with the experience of running those from scratch drivers than what hardware manufactures ship.
Absolutely. A large part of the reason is that in the OSS world the architecture is optimised to make as much as possible common between drivers for different hardware.
For example for GPU drivers Mesa has tons of common code (NIR, GLSL parser etc) that is shared by all drivers with just the hardware specific parts being per driver whereas closed source vendor drivers reinvent the wheel each time.
Similarly for kernel wifi drivers there is a single MAC802.11 stack shared by all drivers.
Vendor drivers have an initial head start since those writing them have access to internal documents describing the hardware interface and don't have to do reverse engineering. But, over time, OSS drivers can be better as improvements to common code help all drivers.
In fact I think the best way hardware vendors could help OSS is not to provide drivers but documentation.
They also don't need to contribute to the Linux kernel. Why would they? They don't support Linux on their hardware in any official capacity, otherwise projects like Asahi wouldn't need to exist.
And playing devil's advocate, Apple has open sourced their macOS and iOS kernels, and has some open source presence[1]. None of their contributions are crucial parts of their business, of course.
[1]: https://opensource.apple.com/
The parent already addressed the point that you're trying to make when he said:
> I don't see how this is any different than early linux when no hardware manufacturers had any interest in supporting it.
This is not early Linux anymore.
I have an AMD Linux laptop I’ve been using for work.
It’s great. The battery life is great, it’s quite fast with a lot of cores, when I need to do my genetics runs (plugged in). Build quality isn’t bad, plus affordable and lots of ports. After my initial transition away, not missing my 2015 Mac book pro.
Linux is the way to go. I don’t blame people with apple hardware for wanting it. I just don’t feel the x86 side is as bad as the everyone makes it out to be. We’ve come along way since my first Linux laptop and it’s not so great battery life.
Two years ago or so I bought a ThinkPad with an AMD Ryzen CPU, there was a lot of hype about them. How Linux laptops were finally competitive, speed, driver, and battery-wise.
The machine was quite a bit slower than an M1 Air, would have loud fans during video meetings, and on Linux the battery would typically last 3 hours (6-7 on Windows, yes I did all the usual power optimizations). In S3 sleep it would discharge overnight and the next day it would refuse to charge with Lenovo’s included USB-C adapter. When waking up the machine from sleep the track point or trackpad wouldn’t come up 1/3rd of the time on Linux.
I used the laptop for work and the question ‘does the laptop work’ when having a meeting or having to teach became so stressful, that one day after another Linux hardware episode I immediately went to a store after work and picked up an M1 Air and never looked back (well, got an M1 Max after that).
There is no way I am going to touch Linux on laptops within 5 years.
(I use a headless Linux GPU machine daily, first used Linux in 1994, and was paid to work on a Linux distribution in the past.)
> Two years ago or so I bought a ThinkPad with an AMD Ryzen CPU
Things have changed a bit since then[1]. The new Phoenix chips are quite competitive with the M2 as far as performance and TDP goes. Your other complaints are with Lenovo, not AMD.
I doubt anyone will argue that Apple laptops don't have the best build quality. Apple has the advantage of full vertical integration, so it's very difficult for any other manufacturer to compete on things like battery life and power efficiency.
The Linux glitches you describe is the usual Linux jank. I don't disagree that even the most well-supported Linux laptop will have these. As a Linux user, you choose to deal with these issues because the alternative of relying on a corporation to decide how you're going to use your computer is not an option. I've also heard and experienced my share of issues with macOS and Windows. In the eternal words of a modern philosopher: every OS sucks[2].
[1]: https://nanoreview.net/en/cpu-compare/apple-m2-max-vs-amd-ry...
[2]: https://www.youtube.com/watch?v=CPRvc2UMeMI
Was the laptop advertised as supporting Linux, or, if not, did you at least do your research ahead of time to ensure that everything worked properly? Because clearly it didn't, so I expect Linux support was already known to not be in a great place before you bought it.
And yes, that sucks. We should have first-class support. It's no wonder macOS gained popularity among developers. But I've been running Linux on laptops for 15+ years now (even on Macs), and I've seen how it's changed from barely-working and having to futz with things at every kernel upgrade, to pretty much seamless (and these days I really have little patience for futzing around with things; I want something that works so I can do useful things on it). But, again: you need to choose your hardware carefully.
For reference, I had a 2016-model Razer Blade Stealth, which had no issues with Linux. Then in early 2019 I bought a 2018-model Dell XPS 13 that worked flawlessly (except for the fingerprint reader, which I knew ahead of time and accepted as ok). For the past yearish I've been using a Framework Laptop, which has had some problems (unrelated to Linux; Windows users have the same problems), but the hardware support on Linux has been solid.
Meanwhile, I'd constantly hear problems from my friends with Macs about how it could never stay connected to a wireless network after a couple hours (requiring a reboot to fix), or would frequently "beachball" under not that much load, or how the yearly major OS update would usually break their development setup. I used macOS on and off between 2005 and 2017 or so, and ran into plenty of issues as well.
While I certainly agree there's some laptop hardware that you just shouldn't run Linux on, the still-kicking Steve Jobs Reality Distortion Field somehow causes people to ignore or explain away all the issues macOS has.
Was the laptop advertised as supporting Linux, or, if not
Yes, it was Linux-certified.
I've seen how it's changed from barely-working and having to futz with things at every kernel upgrade, to pretty much seamless (and these days I really have little patience for futzing around with things; I want something that works so I can do useful things on it)
Dig up any post from 5, 10 or 15 years ago and Linux users will say literally the same thing.
Sorry it didn't work for you. I would recommend anyone buying at notebook to get one with linux pre-installed. I did that because I want to use this thing, not futz with it.
I'm assuming you're using Asahi Linux on your Macs (though you said you wouldn't touch it..). The lack of hardware diversity should make comparability easier, even if everything need to be reversed engineered.
I get 6 hours or so on my machine. Its pretty much silent, unless I push it. Its a Ryzen 7 5700u. We do a lot of parallel compute and genetics code tends massively parallel and x86 optimized. Mostly run on cluster though. I haven't done any maintenance and have had not hardware issues.
I don't link I could ever go back to macos, or windows.
What ryzen laptop is this that you keep referencing?
The negative experiences with the thinkpad are typical of all the intel laptops I have recently used, preloaded OS (including Windows, and to a lesser extent, Linux and MacOS), or not.
Whenever I look for an AMD laptop, it has a low resolution (1080p) display, and/or an off-center keyboard/trackpad (or has some other obvious fatal flaw).
I'm typing this on an M2 macbook. I do 100% of my work in an "8 core" arm Linux VM that can only use one core for userspace stuff (according to top), but that still kicks the pants off my previous laptop.
I'm strongly considering dual booting into asahi.
Its a system76 pangolin (2022)
>Whenever I look for an AMD laptop, it has a low resolution (1080p) display, and/or an off-center keyboard/trackpad (or has some other obvious fatal flaw).
yup it has both of those. But the screen is only15", and I'm old so it doesn't matter. It not glossy which I really like though.
If you love the mac hardware, give Asahi a try. My understanding its the best linux for the M-series macbooks. Linux is great for developing on and they seem to be making great progress.
It's interesting how different tolerances are. I couldn't even stand 1080p on 13.3".
And then it needed some fractional scaling factor. Wayland apps worked ok with that scaling (though rendering was perceivably jankier), but some X11 apps would just be blurry. At the time there was no solution for there apps and looking at an extremely blurry CLion all day is no fun. The only solution was to run the whole desktop environment in 1x scaling and use this GNOME option to use larger fonts and widgets. Which worked ok-ish, but many things are sized in a funny way.
I just couldn't tolerate so much brokenness.
(And don't get me started on sharing a screen in Zoom conference calls.)
I don't understand how scaling is so broken. I switched my Linux desktop environment to a high DPI monitor in 2001 (it was a Sony CRT). Stuff like font and bitmap rendering looked better then than it does now.
The off-center keyboard thing is super irritating. I know some live and die by numpads but for my usage, 99% of the time they're just collecting dust and making typing less comfortable.
Laptop displays are also a common frustration, though this has been improving lately. Still too many models stuck on 16:9 aspect ratio though, which is suboptimal for anything but watching movies due to lack of vertical real estate. By the time you've factored in all the taskbars, titlebars, toolbars, menubars, tab bars, and status bars you've got a keyhole left to peer through. This is less of a problem for those using something like i3 or Sway where half of those bars are hidden but tiling WMs just aren't my thing.
I managed to get Ryzen 5850U/15W Thinkpad with 4k display in 2021. There were some necessary firmware updates during the first year to get the suspend and power management working (afaik, not so many kernel bugs), luckily fwupd now works for Dell and Lenovo so fw updates are easy and there is a popup to recommend them when available. For the past year it has been very stable running Fedora and thus using the latest kernels.
Now waiting for the Ryzen 7840U configured to 15W to be available with 64GB ram.
> I just don't think that it makes up for depending on a small team of volunteers to resolve all hardware issues in an ecosystem hostile to OSS
This. The volunteer pool is too small. And you're supporting a shitty company.
I would like to see the volunteer pool grow, but I suspect a lot of things in the Linux kernel are maintained by a small pool of volunteers--it's not like all volunteers (or paid developers) work across the whole kernel, after all.
Also, which are the virtuous hardware companies and what's stopping them from making compelling products? And since I'm already a Mac user, I've already supported this evil company, so what does it matter if I change out the operating system?
Not sure why you're fixated on the whole Linux kernel - we're talking about a small pool of volunteers supporting complex modules.
There are good laptops out there other than Macs... Lenovo, HP, Dell, etc all have offerings which are supported out of the box by Linux because they aren't using their own hardware or do contribute the necessary code to run FOSS OSs.
I'm not asking you to now dump your Mac. That'd be silly. Continue to use your Mac with Asahi, just don't complain if Apple decides to break anything at any time, and expect it. Know that what you type on now is already planned for obsolescence and most likely has intentional design flaws as shown again and again by people like Louis Rossmann.
When your Mac dies, I am asking you to not buy Apple for your next laptop. That's all that can be reasonably asked.
> There are good laptops out there other than Macs... Lenovo, HP, Dell, etc all have offerings which are supported out of the box by Linux because they aren't using their own hardware or do contribute the necessary code to run FOSS OSs.
For the love of all that is holy, name one model that has the following properties:
- 6 hours real life battery doing C++ development work.
- 7+ days suspend battery life
- 99.99% success rate resuming from suspend under linux (~ 1 kernel panic per year is OK)
- Centered keyboard and trackpad
- >> 1080p screen
- bluetooth, wifi, webcam, etc, etc, all work reliably (~ 1 device "need to reboot" failure across all categories, per year)
- un-noticable fan
- less than 10% permanent hardware failure rate per year
None of the last ~ $10K worth of intel machines I've used (including high end macs, linux and windows machines) met the above criteria.
My M2 macbook actually ticks all the boxes.
However, I really, really, dislike MacOS.
I've managed to tick all your boxes on not one but two Lenovo ThinkPads. Ironically I've had trouble with suspend on my previous work Mac laptop running macOS. Also issues with WiFi on the same Mac laptop not present on other devices. Every single hardware failure within warranty I've had on both personal and work laptops have been Apple - LCD, battery, mainboard, dvd drive, just to name a few. Yet all the others I've owned have done their three year duty without fail. Very much a case of YMMV.
The end of the intel mac laptop line models were the worst laptops I've ever used. My last one went through 2 screens, the battery swelled, it couldn't charge fast enough to stay on while plugged in (and the CPU ran at 10% perf while charging, but only if plugged in on the left side).
Also, at any given time 6 of the keyboard buttons stuck, and the touchbar constantly phantom pressed the siri button when I pressed backspace. It got under 90 minutes of battery when brand new. It reliably kernel panicked on resume 1-2 times a week. It was also loud and hot.
Maybe I should give lenovo another try.
They completely screwed me over on a warranty repair right after the IBM acquisition. When I got the laptop back after over a month, it was diagnosed "no fault detected" and had a new symptom: it leaked high voltage from the backlight transformer into to the case during boot shocking the heck out of me!!!
Also, when I look at their web page, I always have the problem that they offer too many sub-configurations, and there isn't a button that says: "just give me the one that definitely runs Linux with the taint bit turned off, and only contains components that have had stable OSS drivers for over a year, and whose BOM hasn't changed for at least 12 months".
Ironically, Apple's web page is pretty close to having a de facto button like that.
My thinkpad does all those things... I expected really something crazier as a rebuttal. It sounds like you may have bought bad products and then bought a Mac when they had those specs and were happy then...
Really? My 2016 ThinkPad Carbon hasn’t been able to run off of a charger since ~2019.
To be clear, is this a description of your experience with Asahi Linux?
No; with MacOS running an Ubuntu Multipass VM. :-(
I wish more manufacturers worked harder to keep fan noise down to between none and barely audible without severely throttling the CPU and GPU or toasting your lap. My laptop shouldn't be spinning up its fan just because I plugged in something as pedestrian as a 2560x1440 60hz monitor (as my ThinkPad X1 Nano likes to).
Installing "uBar" fixed a lot of my beef with MacOS. (I get no referral bonus, just a happy user)
[flagged]
Aren't lots of things in Linux dependent on a small team of volunteers? I know the Linux foundation owns the whole kernel, but in practice how many full time people work on the ext4 driver or whathaveyou?
/me glances at their full NixOS desktop that native builds for x86_64-linux and aarch64-linux. Heck, most all of it cross-compiles too. Native riscv64-linux needs more attention (upstream support is missing in places) but most of my config cross-compiles fine as well.
Every month that passes, with similar HN comments insisting it's a bad time... I wonder: "am I 'special' or just spoiled by nix(os/pks)". Or maybe, just maybe, people's expectations of their distros are shockingly low. And maybe rightly so at times.
> I'm sticking with my slow, hot and power-hungry x86 machines with worse build quality for the foreseeable future.
Nothing wrong with being a late adopter. Nothing wrong with being an early adopter either.
> And the incompatibilities with ARM are not negligible.
Namely?
The compute accelerator story on mainstream, non-patched Linux, with upstream software isn't that good at the moment. You're going to be waiting a while before you can do fun stuff like organize layers across the Neural Engine and GPU for ML models, something CoreML can do today. Compute using graphics APIs exists, but it isn't really the same and loses out on many features people practically want and are used to, and it moves forward much more quickly than graphics APIs e.g. Nvidia just released Heterogeneous Memory Management as stable in the open source GPU driver for x86. The Linux accelerator ecosystem in practice is just held together by Nvidia's effort, honestly.
We really need something like Mesa, but for compute accelerator APIs. I'm really hoping that IREE helps smooth out parts of the software stack and can fill in part of this, but the pieces aren't all put in place yet. You'll need the GPU for a substantial amount of accelerator work regardless of Neural Engine support.
I disagree that there is nothing lacking on these machines with Asahi, I still run into small nits all the time (from 16k page sizes biting back to software missing features). But my M2 Air is 100%, no-questions-asked usable as a daily driver and on-the-go hacking machine, it is fast as hell and quiet, it has nested virtualization and is the only modern ARM machine on the market, and I love it for that.
Is the Neural Engine/CoreML used in "normal" desktop apps on macOS, or is it limited to specialized ML centered apps? In other words, where should I expect performance improvements if there was a hypothetical Mesa for compute accelerators? Spontaneously, I can only think of image editors like Photoshop offering AI-based tools.
The features support page lists the webcam as TBA. How do you do video conferencing? USB webcam? No video?
People who demand others show their faces in a video call generally aren't fun to be around anyway, and if it's family or friends just use your phone.
Typical HN response "This major feature most people use doesn't work" "Just don't use that feature, it wasn't that good anyway"
“Sorry, grandma, I can’t video call with you. I’m a Linux user.”
> when the neural engine drivers come online now
Has there been any ongoing work on this? It's been marked as "WIP" in GitHub for a while now, and I'd imagine it's one of the more complicated things to reverse-engineer.
You can check here for some insight
https://oftc.irclog.whitequark.org/asahi/search?q=ane https://oftc.irclog.whitequark.org/asahi/search?q=neural https://oftc.irclog.whitequark.org/asahi-dev/search?q=ane
Some: https://news.ycombinator.com/item?id=35301630
Would you say buying e.g. a Mac mini for 2.3k€ just to run Asahi Linux is worth it?
For €2300 I assume you're looking at an M2 Pro model? Note that neither the M2 or M2 Pro Mac Mini currently have working display outputs[1] so no you should not. Apple changed the way the display outputs work in M2 so they're now dependant on Thunderbolt/DP alt mode support which is not implemented for any Apple silicon machine yet.
On the other hand a cheap M1 Mac Mini would make a great machine to try it out. The M1 Mac Mini is the best supported machine currently.
[1] https://github.com/AsahiLinux/docs/wiki/Feature-Support#m2-d...
I don’t run Asahi on anything currently, but I do have two desktop Linux machines, an M1 Macbook, and have previously run Linux on an Intel Mac… I can see the argument for laptops based on battery life/heat/build quality, but for a desktop machine I’d need a lot of convincing to justify the price premium and risk of compatibility issues in choosing a Mac Mini over a SFF/USFF/Tiny desktop with fully supported hardware.
I’ve owned both windows and Apple computers, quite many of them, over the last 20 years. On average, the Apple machines last at least twice as long as the windows machines while still being fully usable. One could argue just based on that basic math that they are worth twice the price.
Compared to what? Junk? My w541 is 10 years old and I just ordered parts from Lenovo to perform cosmetic repairs on it.
If you take care of your devices they will last. - typed out on a gen 1 i7 desktop
Unfortunately, Apple machines are usually 4 - 10 times more expensive, making this choice still quite difficult.
They are expensive but 10X certainly seems like a stretch. Show me comparably specd hardware only 10% the price of an Apple machine?
See, here's where the undefined nature of things comes in. "Comparably spec'd" needs to be conditioned on what task you're aiming for.
A "pure gold hammer" is a terrible idea and would also be terribly expensive. But asking for a "comparably spec'd" hammer presumes the absurd premise that the material of the hammer must be kept consistent regardless of its intended use just for the purpose of being comparable.
To preface, I totally understand the value proposition of Apple devices for some use cases, but it is important to recognize that they are aiming for certain workloads.
As examples:
I have one friend that runs virtualization workloads that require a lot of memory, a lot of storage, a lot of cores, but they don't really care about memory bandwidth, "having a display", or even the noise of the device. An older server with 192G of RAM, 24 cores and >8TB of storage can easily be had and upgraded within $1k, whereas a "comparable" Mac Pro costs upwards of $10k! (Of course nobody would use a Mac Pro for this workload, so the comparison is moot)
I also have friends that are digital artists. They care about having a high brightness and color accuracy display but otherwise don't do anything that taxes the computer. They also appreciate having high quality speakers and long battery life. Some of them run M1 Macbook Airs at the lowest 8G memory configuration for ~$800 (discounted new from other retailers) + a digitizer for ~$100, while the closest comparable non-Apple laptops are all premium devices upwards of $1.5k and even then they are still worse in the battery department!
As for myself, I do light dev work, virtualization, gaming, but also travel a lot and present at conferences. I use a GPD Win Max 2 for a little over $1k (early Indiegogo pricing). The closest Apple offering would be a 14" MBP, and configured as needed (32GB/2T) would be about $3800 and still be short a 4G modem and a couple of extra devices like a digitizer, game controller, and dongle for USB-A. ¯\_(ツ)_/¯ Can't win 'em all.
> An older server with 192G of RAM, 24 cores and >8TB of storage can easily be had and upgraded within $1k,
Are you referring to a used server, or just buying a minimally-equipped new server and upgrading it with aftermarket RAM and (low-quality) SSDs?
Used (decommissioned from equipment retirement from companies) server, upgraded by maxing out the RAM slots and using the cheapest available SSDs.
This is a pretty common practice for homelab enthusiasts, or so I hear.
A micro ryzen 5600U build with really bad quality components can be half the price of a mac mini with equal CPU performance. If you upgrade all of the mac specs you can probably get a larger divide, but IMO maxed out macs don't make much sense for most people.
It's that focusing on one specific aspect of the system and compromising on everything else that produces the really big discrepancies. I tried to use PCPartPicker to spec out a rough equivalent of a maxed-out Mac Pro in terms of CPU cores, GPU performance, and RAM and SSD capacity, but still ended up at with at most a 3.5x disparity, and that's ignoring the GPU VRAM capacity limitation and features like Thunderbolt and 10GbE and assembly and warranty and support. If you want to assign $0 value to a large portion of a Mac's features then you can make it look wildly overpriced, but that's mostly an admission that it's the wrong product for your needs.
The years of the keyboard issues left a bad taste in my mouth, but I switched to a non mac laptop for my previous laptop and now I'm back again. The coupling of the OS and hardware really do make for a great user experience. I don't want to play games on my laptop, which is the only real use case where I hear valid complaints. I just need my dev environment and snappy research and communication.
A valid complaint from me is linux based container resource utilization. The only really good fix for that IMO is if apple did something like WSL2 or FreeBSD's linux ABI and had an efficient compatibility layer. For now I just run dev containers on my (linux) desktop.
A 2X price difference is certainly believable, but I was responding to the suggestion of a 10X price difference.
You mean it’s possible to buy the exact equivalent of a M1 MacBook Air for £99-£249?
Send me the link, that sounds amazing.
Yeah, the M1 MBA was really so out of the norm in terms of value that it's pretty much impossible to find anything like it at the price point. It literally turned me into an Apple person, essentially overnight. I no longer even switch on my windows machines.
Not sure where you got the impression of that?
There do not exist "equivalent"s to any Apple devices (I don't see them licensing the M1/2 chips to anyone else anytime soon). But depending on what you care about, a "comparable" Apple device could be 10x more expensive. Of course, for other tasks a "comparable" Apple device can also be _cheaper_ than any non-Apple device available!
Only looking at aiming for similar "longevity" (since the parent is using that as a benchmark), there are plenty of devices that have a useful life comparable to Apple devices at 1/4 - 1/10 the price.
I'd love to see your 10x example.
There is one above in the server/homelab space. Items like memory and storage are charged huge markups* so if you need a lot of capacity of those you are going to quickly get into the 10x range!
As for longevity, if you consider software support ending as EoL, software/OS support for a huge swath of Intel iMacs (especially those with DGPUs) was dropped by Apple quite a few OS releases ago and you have to run community patches to keep them working. Whereas similar decade old hardware still run Win 10 and Linux out of the box.
*: Don't get me wrong though, the markups are for good reason. x86 platforms don't offer anything close to Apple's ARM chip memory bandwidth (which are closer to DGPU levels). Similarly, for flash/SSDs.
My M1 MacBook Air (8GB RAM) was cheaper than the 16GB Dell Latitude 13 i7 I bought a year and a half previously. I am much happier with my MBA, even at 8GB, than I ever was with my Dell.
You can get a used M1 mini for more or less 400€. Get a glimpse of whats going on in your local facebook marketplace, most likely you'll come up with nice offers.
I'm guessing that's for a model with 8G memory?
In my experience the experience for those is quite bad, as you're sharing that 8G across both the CPU and GPU...
Judging from the OP's post of 2.3k€, they're probably considering a maxed out version, which has a completely different experience since you can fully take advantage of the high memory bandwidth for hybrid tasks unlike the low-memory models where you're sharing the limited capacity.
You must be kidding right? Who on earth would sell for 400?
Literally the first result on (US) eBay is for 419 euros with "more than 10 available"
The Mac mini M1 was on sale new for $400 at Costco a couple weeks ago. The M2 Mac mini is $499 on the Apple education store.
Not the OP but I got a 13" and a larger model Mac of the x86 variety when they were still reasonably young and even though I eventually got all of the bits and pieces to work it usually pays off to wait until a somewhat larger distro supports the hardware as well. That way you benefit from a much larger crowd of testers and once they have no more issues you should be good to go.
Moneywise it was definitely worth it, both machines are still working many years later and have been pretty much trouble free after the initial bugs were ironed out.
If I was in the market for a new laptop right now I'd wait for a bit and then pull the trigger on the latest model with broad support.
I'd say getting a macbook or a macbook air would be worth it, but rather than spending that much on a mac mini I'd probably get one of those new ryzen mini-PCs like from Beelink or Minisforum. You could get something with a 7735HS 8-core chip, terabytes of diskspace and a shitload of LPDDR5 RAM for 500€ and it's as small as the mac mini.
New M2 mac minis start at $600 (8 core CPU; 8GB ram; 256GB SSD). You can probably find similarly specced x86 PCs for a comparable price, but this doesn't seem unreasonable. https://www.apple.com/shop/buy-mac/mac-mini
I'm running it on my M1 MBP. Also super happy. How do you use suspend?
That’s the only thing I’m really missing currently.
Curious, what do you use it for?
[dead]
Donate.
Please don't forget to donate if you get value from Asahi.
This is tremendously detailed and laborious work that people are doing in their free time.
https://asahilinux.org/support/
Why does this posting have a rel=noreferrer tag, when others don’t?
Looks like jwz links also have a "noreferrer" tag added as well.
https://news.ycombinator.com/item?id=3132752
New silent HN policy to avoid showing its users that some people don't like them.
This is extremely gross 'Dang, and I'm wondering how you can possibly defend doing this. Seriously, why are you going to these lengths to evade a soft-block on links to specific sites that very clearly and explicitly are communicating that they don't want HN's traffic?
Can’t agree more. Disgusting.
the moderation here literally doesn't care about the authors and developers who provide the links to the articles.
routinely they let them be harassed and derided on HN and do not care to focus on their attempts to stay off and away from that.
it's completely expected but still shameful.
Yeeeeeah, your stretching with your definition of harassment.
adding a config to your webserver to evade a soft block is directly enabling continued harassment. Its not passive, not "I'm just getting out of the way here", not "I can't stop this, not my fault", you are literally editing your webservers config file for a site-specific change to allow your forum visitors to evade an explicit block, a block enacted because of harassment.
I mean that's because you consider it harassment. But if you consider it simple criticisms, suddenly its the person not accepting them, not harassment.
Celebrities deal with much worse, yet they don't consider "Taylor switft is ugly" harassment.
it*
Yeah this is pretty inexcusable.
That page doesn't include noreferrer on the article link for me.
"rel" is an attribute of several HTML tags, including the anchor tag <a> [0]. It is not something that would be visible in the URL.
The presence of this attribute instructs the client to not send the HTTP header "Referer" [1] when making a request to the target URL.
[0] https://developer.mozilla.org/en-US/docs/Web/HTML/Attributes... [1] https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/Re...
I know, I'm saying that it's not in the `rel=` on that article link on the linked page. Just noopener, which is of course included on every link.
It is. When I open the source, I see this:
<a href="http://www.jwz.org/blog/2011/10/eff-declares-premature-victo..." rel="noreferrer">Jwz: Google's pseudonym support "obvious bullshit"</a>
Might be that the method is different on a specific browser, but most likely you are looking in the wrong place.
<a href="http://www.jwz.org/blog/2011/10/eff-declares-premature-victo..." rel="noopener" target="_blank">Jwz: Google's pseudonym support "obvious bullshit"</a>
> Just noopener, which is of course included on every link.
I can see no `rel` attribute on other submitted links.
Are you perhaps using some privacy-enhancing browser extension which adds the `rel=noopener` attribute to every link?
Something related to this? https://social.treehouse.systems/@marcan/110503050993279759
I wonder if the new Mac Pro's full PCI Express support resolves any limitations that prevents people from using GPUs over Thunderbolt on existing Apple Silicon hardware (this is apparently a hardware limitation).
Although the Mac Pro's PCIe extensibility makes it a pretty mystifying niche product from Apple without providing memory and GPU expandability, once Asahi Linux gets running on there you should be able to not get the full abilities of the latest Vulkan and full OpenGL 4.6 by putting in a recent AMD card. The open source Radeon drivers should "just work" on ARM as they do in the Talos II POWER-based workstation, if they can be stably initialized that is. Heck, Nvidia publishes a binary Linux aarch64 driver and they sound petty enough with Apple to try to make that work.
You could have Asahi Linux running and delegate any not-yet-supported hardware to the 7 PCIe devices it supports. Would be quite a mighty ARM Linux workstation. Again though - only if Apple has the PCI Express support for it.
@Marcan shared some technical details around their PCI Express implementation recently here: https://social.treehouse.systems/@marcan/110494017883893557
It seems they didn't make any massive changes and instead just put switches on the existing PCI-E Lanes. That probably doesn't bode well for full GPU support :(
There are no GPU drivers anyway. So I’m not sure it matters (Linux or macOS).
Seems designed more to support lots of low/medium bandwidth cards, not a few high bandwidth cards.
This is great work and I commend it. But in other threads people are acting like Asahi Linux hardware support is 100% complete. My fear is that if I were to go this route and purchase the hardware I'd be seeing fraction of the performance and capability I would in Mac OS. To be honest this blog post seems like the project has a long ways to go, not that it is nearly completion.
I just can't justify buying hardware from a company that is so hostile to developers and hackers as nice as it may be.
> I just can't justify buying hardware from a company that is so hostile to developers and hackers as nice as it may be.
I don't think it's hostile, I think they're just hands-off; they throw the hardware over the fence and say, "if you wanna make use of it, here's our software; if you don't like our software, sorry no docs but you're free to write your own". Which is exactly what's happening.
I mean it would be nice if Apple had released more documentation, but I totally understand if they don't want the burden of supporting it.
First, personally I don't care what hardware or software people use, if they are happy with the tools that they using then that's good.
That said, Apple has been very hostile to hackers over the years imo. Hardware being hard to repair, access, upgrade, etc. I think at one point they were making it virtually impossible to replace components because they were serial locked.
As far as I am aware, progress Apple as made has been in response to public image issues or changes in consumer laws within the EU.
Plus Apple software is heavily indebted to Open Source software so they very easily could be releasing drivers for their hardware instead of relying on community to do backwards engineering.
> Hardware being hard to repair, access, upgrade, etc. I think at one point they were making it virtually impossible to replace components because they were serial locked
You can only have so many flexibility in design with modern hardware — they are not fitting things into 5 cm “thin” chassics anymore. How exactly are such a thin device be repairable? Similarly to how old car motors could be tweaked with, you need special tools to touch anything in a modern engine. This is not against the customers, these are trade offs.
But even this way, apple devices have by far the longest lifetimes, macs, iphones will have 2-3 owners easily - so is it really fair to call them out, or is it just baseless emotional reaction?
Also, what you heard about locked down components resulted in better security, a much lower risk of theft, and a much more clean second-hand market (where you won’t be sold a phone with a cheap chinese shittier screen for example).
> But even this way, apple devices have by far the longest lifetimes, macs, iphones will have 2-3 owners easily - so is it really fair to call them out, or is it just baseless emotional reaction?
I would dispute this claim, e.g. Apple settles iPhone slowdown case for $500m[1], just the first link I found looking for "planned obsolescence apple" on DuckDuckGo. This is not exclusive to their iPhones as one can find with a quick search.
[1][https://www.bbc.com/news/technology-51706635]
Which was about a very stupidly communicated, but genuinely good intentioned misshap. Iphones older devices with old batteries had cases where they no longer could provide sufficient voltage to the phone and thus it would reboot randomly for people. To fix the issue, they decreased CPU frequency so that it would no longer drain so high peak power.
But they should have made it an option (it is one now on these devices) instead, they might have even come out good from it (as an android manufacturer wouldn’t even care about such an old device at the time).
"But even this way, apple devices have by far the longest lifetimes, macs, iphones will have 2-3 owners easily - so is it really fair to call them out, or is it just baseless emotional reaction?"
Why reply that criticism of Apple must be purely an emotional one? Kind of diminishes your argument here.
Immediate search result for repairable phones:
https://www.androidcentral.com/best-sustainable-repairable-p...
https://shop.fairphone.com/en/buy-fairphone-4
Here's a laptop that you can upgrade:
https://frame.work/
Lifetime for Apple isn't as long as you make it out to be when batteries need replaced and software support for hardware ends:
https://arstechnica.com/gadgets/2023/06/macos-sonoma-drops-s...
"Also, what you heard about locked down components resulted in better security, a much lower risk of theft, and a much more clean second-hand market (where you won’t be sold a phone with a cheap chinese shittier screen for example)."
Apple could just release stuff that didn't break so easily too so no need to risk changing out a screen if it ain't broke. There are plenty of ways to increase security of the device without making it less consumer friendly.
Additionally since the context here is whether Apple has been hacker friendly or not, why shouldn't you be allowed to upgrade and change the hardware of YOUR device? As in, you want to put in more storage or change the screen to one that's better in some manner (maybe it's just cost) then you ought to be able to.
That is it should be the device owners choice whether or not to replace their screen with one from Apple or a cheaper one.
Going with the first link. The "best" phone is not globally available. The second "best" comes with three years of OS updates. I stopped reading there; three years is a much, much shorter lifespan than you get with iOS or MacOS.
Also, say you have one of these phones, and are in a major city, then break it. How will you get the parts you need to repair it? How many hours will you be without a phone?
With Apple phones, it's typically same day service to get it repaired. Worst case, you can get a new phone with your data mostly transferred, again, same day.
The Ars article you link is pointing out that Apple is dropping software support for laptops that are 6 years old. That's better than pretty much any other vendor.
As far as laptop repairs go, frame.work is probably the best non-apple option, but they don't have a fixed policy for how long replacement parts will be available. The story is similar for Apple:
https://support.apple.com/en-us/HT201624
says they provide parts for up to 7 years, and battery swaps for up to 10 (subject to part availability). I hope frame.work will be able to do better, but I challenge you to find any laptop company significantly better than this.
(Other than soldered ram and disk, I honestly don't care about third party parts. It's not like Apple replacement part markups are insane or there are significantly better parts available. I've definitely never used third party parts for other brands of laptops, even when they were available. However, I've been repeatedly screwed over getting other brands of laptops repaired, especially under warranty.)
Anyway, I get why Apple has a bad reputation for support and repairability. There objectively bad. However, that doesn't mean they're not simultaneously also the best option (or close to being the best).
I am sorry but you're missing the point here.
I'm not arguing over better hardware (performance) or price. I am arguing over hack-ability; that's it. I hope you can understand that.
None of the machines on that list are more hackable than a Mac though.
Some are more "hackable" than an iPhone, but only in some strange symbolic way, since once you de-googled android is somehow more user-hostile than iOS and completely unhackable in practice (in the sense that I can't make it do what I want and also be usable as a daily driver that lets me do stuff like pay for stuff, use public transit, charge my car, park, or take an uber/lyft).
I'm a happy owner of a pine book pro, and a pc engines router; I get it. However, I don't think there are any viable Linux laptops or phones that compare favorably to the linux laptop + phone I had a decade ago.
Scratches head at how framework laptop is less 'hackable' than a Mac but okay.
"(in the sense that I can't make it do what I want and also be usable as a daily driver that lets me do stuff like pay for stuff, use public transit, charge my car, park, or take an uber/lyft)."
So either the hardware you bought isn't letting you do what you want with it (this includes iPhone too!) Or you become a better hacker and get it to work yourself.
> I'm not arguing over better hardware (performance) or price
Neither was the comment you replied to.
They were arguing that Apple provides excellent hardware support (in terms of repairs and how long they take) and lifetime (in terms of how long updates are provided for the device).
Hackability is irrelevant. Apple clearly doesn’t target that market and hadn’t since the Apple II days.
Would you buy a Lamborghini and complain it makes a terrible truck?
That’s not something Apple designs their products for, and their users are either fine with that or willing to make the trade off for the other things Apple does prioritize.
So the only option has 1/10th the hardware of an older iphone with shittier camera than that.. for me that’s not a good deal. Framework laptops are insanely expensive for again, worse hardware. Isn’t the point of linux to run on anything?
Also, every device needs battery replacements, like this is just the physics/chemistry of batteries and it has an absolutely doable price for any apple device.
What breaks easily on an iphone? They are quite sturdy phones with metal casing. They wouldn’t get sold after 5-7 years of active use if they weren’t sturdy. And glass will still be breaking when it meets with big enough force - I again don’t see your point.
> re hacker friendliness
The RAM has different architecture on the M series, so it can’t be replaced even theoretically. Also, every moving part is one more point of potential breaking, plus it takes up space. This is not a rasppi, different design goals/constraints.
You can put in a worse screen but one will be able to see that in the settings so they can’t be scammed.
> Framework laptops are insanely expensive for again, worse hardware.
1199 EUR is not insanely expensive, specially considering that I can put up-to 64GB of RAM in a Framework laptop with a reasonable amount of money, while I would need to pay almost the price of a full Framework laptop to do the same in a Macbook Pro [1]. This is IMO, insanely expensive.
And yes, I can definitely use those amount of memory during mass rebuilds that I sometime like to do in NixOS. I don't even try to do those same workloads in my macOS because they start to become hugely slow once you hit the swap.
[1]: by the way, this get even worse considering that I also need to upgrade from an M2 Pro to M2 Max to have the option to do so. I just did a quote for the cheapest Macbook Pro with 64GB of RAM, and I got a 4000EUR quote for 512GB of storage that is laughable low for something that expensive. At that price, I can get 2 of the most expensive Framework AMD and I would still have sufficient money to get another one of the older Intel ones as a spare.
Mind you, the two kinds of RAM are not directly comparable.
Sure, I never said so. It doesn't mean that there isn't workloads that benefit from more RAM (even if it is slower) or that the Apple's RAM prices are insane.
Not to mention I can't find anything on them having ECC capabilities but I could be wrong.
If you like the hardware or software that you use then I am happy for you.
"This is not a rasppi, different design goals/constraints."
That we can agree on.
Apple is a business with specific goals and so far as a business in terms of profits they have been successful.
All I wanted to point out was that Apple is not hacker friendly in my opinion, and I have listed good reasons that you don't want to accept. There's no amount of going in circles here that will change either point of view I fear.
Have a blessed day!
> That said, Apple has been very hostile to hackers over the years imo. Hardware being hard to repair, access, upgrade, etc. I think at one point they were making it virtually impossible to replace components because they were serial locked.
That came partially out of the desire to reduce the lure for thieves and robbers. It was really bad during the first generations that regularly had jailbreaks and ways to bypass "Find My..." or whatever, then the first tightening reduced resale values of stolen iPhones by a good amount (as they were only good enough to slaughter for parts once reported stolen), and the latest round made it even worse for criminals.
Personally though, I'd preferred they simply provided "unlock codes" with a phone that could be used to remove the association between a part's SN and the IMEI/SN of the phone. That way, buyers of iPhone have something similar to a certificate of authenticity.
“I think at one point they were making it virtually impossible to replace components because they were serial locked.”
They are very much still doing that
In fairness, most instances of them doing that actually significantly increase the cost of evil maid hardware tampering attacks.
If I could, I'd configure grub or whatever to serial-lock my Linux install to my desktop hardware (and keep a recovery key that would unlock it at another location).
The issue is that Apple isn’t giving anyone access to the tools to pair the parts, unless you give them all the information in advance, buy them at possibly inflated prices through their self repair program if they’re even available, and then have Apple remotely approve it afterwards (and this process only really works for individuals, 3rd party repair is more important as most people don’t have the skill)
If I was running a third-party unauthorized repair shop, then I'd care. However, I just don't see why I should care, as a customer. Also, I have trouble wrapping my head around the standpoint that people want to pay the Apple "it always sort of works" tax, then spend lots of extra time trying to save $50 on a repair.
When I mess around with hobby hardware, it's esoteric stuff, not readily-available laptops that are being built by the millions. For that, sometimes you need a volt-ohm meter, or diagnostic rigs.
Every few years I even bring some piece of a Linux box to a repair shop, to narrow down some fault.
However, Macs are all identical, so repairs boil down to "yank component, throw in the recycling shredder box, and (if you just replaced the mainboard) restore from iCloud Backup". Apple hardware switched over to being livestock, not pets a long time ago, so I'm not seeing the point in spending lots of effort on custom repairs (vs. replacing + recycling the bad components).
Because as a customer you should have the choice to get a repair from whoever you want. Even if you only go to Apple, competition will force their prices down somewhat.
"pay the Apple"
It's not just Apple doing this
"then spend lots of extra time trying to save $50 on a repair."
What if you're trying to save precious data? Apple don't make much effort to do data recovery because they don't care; you can fund 3rd party repairers who absolutely do. What if you got the device second hand? What if it's not $50, but $1000+? I've seen a video where Apple tells a journalist they have to buy a new Mac, and a 3rd party repair shop fixes it for free because it was such a small issue that Apple didn't even check for. https://www.youtube.com/watch?v=o2_SZ4tfLns
"Macs are all identical, so repairs boil down"
This is the problem; Apple is pairing parts to devices, so if you use a 100% genuine part from a new device it still won't work
> What if you're trying to save precious data? Apple don't make much effort to do data recovery because they don't care; you can fund 3rd party repairers who absolutely do
I keep backups, both in iCloud and elsewhere. Apple makes it a headache to avoid doing this, and they apparently finally added proper E2E encryption support.
I guess I think of Apple computers more like generic corporate fleet vehicles, and Linux boxes more like classic cars.
I really don’t think anyone cares if a ford dealer irreparably destroys 1% of the bottom trim white F-150s they maintain, and also forbids the people that bought the service contract from opening the hood.
None of that applies to a mustang shelby, but that’s a different part of the market.
Maybe some people really like their macs / phones, and don’t just use them like interchangeable boring tools.
> If I could, I'd configure grub or whatever to serial-lock my Linux install to my desktop hardware (and keep a recovery key that would unlock it at another location).
This is the general idea behind TPM/Secure Boot, but as you present it, it just sounds like a headache for performing system recovery, at no obvious benefit for security.
What's your threat model? In the 99.(9)% case it's a crook snatching the laptop, wiping the HD, and selling the whole thing and/or the parts. Evil maid is a real threat, but only practical (in terms of sophistication/cost vs benefit) for high-value targets, like C-levels, devs holding company secret keys, etc.
I don't really care if the desktop gets snatched. I'm more worried about old tax returns, the credentials stored in its password manager, etc. Breaking in twice (once to install a bump in the wire key logger, and once to walk away with the machine) would lead to at least 10x more payout for a burglar.
Also, ignoring what it is worth to the attacker, having to roll over all my credentials, freeze accounts, etc, etc, because my desktop was stolen would cost way more of my time than buying a new desktop (happily, the drive is encrypted).
Sounds like paranoia.
> I think at one point they were making it virtually impossible to replace components because they were serial locked.
That seems like an obvious reaction to the fact that criminals (once they were prevented by activation lock from selling stolen iPhones) started parting out stolen iPhones and selling the parts to repair shops.
I think a better approach would be for Apple to only block the replacement part if that part has a serial associated with an activation-locked phone, but I'm sure it's easier to just block everything except for the replacements shipped directly from Apple.
That's not true for Macs. Apple allowed and worked with MS to allow windows to work on Intel macs. That's pretty insane from allowing people to do what they want with their hardware.
No offense but that's not even contextually the same thing so I'm not really sure what you're trying to say.
You claim Apples Mac line of products is adversarial to hackers. Which I dispute with the example they help alternative OS to run on the Mac. First with Windows and now Asahi.
There's a line somewhere, and I think it's different for every person. For a lot of people, the pricing itself is just outright hostile. 8gb of RAM in a base model Mac is not future-proofed in an age of local AI models, and paying to upgrade it gets expensive, fast. For others the OS is hostile when it unceremoniously drops support for $THING they use, or because it has the gall to show them ads. For others yet, the hardware is hostile because it's basically a black box that Apple withholds documenting to maintain a monopoly on fixing them.
Apple is one of the few companies smart enough to deliberately do this. It is both a testament to ability to do brilliant things, and akin to being trapped in a room with a lion that has twice your SAT score. The "golden handcuffs", as they say.
There is literally no other laptop on the market that would be anywhere close to an M1 air in terms of performance-battery life combo. And it is definitely not even on the upper end in price for that, tiny windows laptops are ridiculously expensive!
So, are they really expensive?
Well overall if you sum up what even a basic m1 mb air packs, be it the panel, quality peripherals and soc itself its rather reasonable and within reach for almost anyone contemplating of buying a new computer. When you factor in that they were the first ones with an actual capable desktop arm offer I'd even say its cheap.
At least this is how I felt 2 years ago back when I bought it and more or less even to this day. I'm wondering how is the market as of now, two years later, and how the $1k arm laptops coming out today compare.
Macbooks are really cheap for what they are, not sure what are you on friend. I would never buy one though, still, they are very impressive and its a shame that they are not free.
>My fear is that if I were to go this route and purchase the hardware I'd be seeing fraction of the performance and capability I would in Mac OS.
The performance is there, it has been running stuff much faster than the vast majority of Intel/AMD laptops for over a year.
Regarding the capabilities not sure which one you miss. Do you plan to use it for development, or you want some kind of gaming/multimedia setup?
>To be honest this blog post seems like the project has a long ways to go, not that it is nearly completion.
It's the other way around. It has been usable as a daily driver for ages.
>I just can't justify buying hardware from a company that is so hostile to developers and hackers as nice as it may be.
Then don't?
> Regarding the capabilities not sure which one you miss. Do you plan to use it for development, or you want some kind of gaming/multimedia setup?
> It's the other way around. It has been usable as a daily driver for ages.
Honest questions since I haven't been paying attention to Asahi for some time now:
- Does hardware accelerated video decoding work? Including in Firefox?
- Does sleep work properly or do I get significant battery drain after leaving it sleeping during the night time? Also, does it wake up from sleep reliably? Like if you open/close the lid 100 times in a row would it crash?
- How is wifi? Does it work as fast and reliably on Linux as the Intel cards? Supports latest WiFi standard and 6ghz?
This would be my most basic questions to buy MacBook as a daily-driver Linux laptop.
I don't see why it is unreasonable to desire Apple to provide documentation and open up their hardware. Honestly I'd ask the same as any other hardware vendor. Intel for example provides very detailed technical documents on their new GPUs (A770).
> Intel for example provides very detailed technical documents on their new GPUs
Intel sells parts to other people, and those buyers expect that kind of documentation. And when it comes to GPUs they’re in 3rd place for high end GPUs and are incentivized to give people as much info as possible to optimize for their hardware.
Apple doesn’t sell their computer hardware to any other OEMs, or hsve a 3rd party making their OS, so they don’t need to produce documentation for anyone.
I agree it would be nice if it was documented. But absent a law forcing them I don’t see it happening.
It's not unreasonable, it's just not gonna happen any time soon.
You don't create a new bootloader that allows users the freedom to run an unsigned third party OS without having it degrade the system's security if and when they boot the native OS because you are "hostile to developers and hackers".
My laptop can use TPM and a hardware key with my keys and free software. Where is the degraded security?
I could trick you into adding a hardware key, then install a tampered version of Windows with it.
(Also, the last time I looked, TPM keys could be grabbed with ~ $100 of hardware, but I think that's fixed by some newer standard.)
But, yeah, it's not a big tradeoff in practice. I think their point was that Apple had to expend effort to enable the use case, which isn't "hostile" toward the use case.
> I could trick you into adding a hardware key, then install a tampered version of Windows with it.
There are other issues as well.
For instance, on a PC the security settings are applied per machine and not per partition, so you can't mix an unsigned OS on one partition with full security on another partition.
Also:
> On Wednesday, researchers at security firm ESET presented a deep-dive analysis of the world’s first in-the-wild UEFI bootkit that bypasses Secure Boot on fully updated UEFI systems running fully updated versions of Windows 10 and 11.
Despite Microsoft releasing new patched software, the vulnerable signed binaries have yet to be added to the UEFI revocation list that flags boot files that should no longer be trusted.
https://arstechnica.com/information-technology/2023/03/unkil...
> For instance, on a PC the security settings are applied per machine and not per partition, so you can't mix an unsigned OS on one partition with full security on another partition.
This wasn't true for a low-end Acer I bought a while ago, and it's not true on an Asus motherboard I use. You can add keys to the bios, and then it'll let you run with either key. That lets you use the grub shim key. On the Acer, you can even tell it to screw PKI, and just check that the hash of the bootloader hasn't changed.
You seem unfamiliar with the Alder Lake compromise.
Indeed. Care to give a link? Quick DuckDuckGo search returned nothing.
Sure: https://arstechnica.com/information-technology/2023/05/leak-...
The gaggle of moving parts that are involved in the PC world make security and privacy substantially more challenging because of nonsense like this - a vendor with rubbish security (not even an HSM for critical signing keys!) compromising the broader world.
Thanks. It seems this confirms that my own keys are more secure, because with them such problem couldn't occur.
> I'd be seeing fraction of the performance and capability
You'd temporarily lose some hardware support (documented) while it's being worked on. But I'm not sure why you expect losing performance? This is running native code. Same binary will run the same on both systems (+/- the llvm version differences).
Does the CPU run at similar frequencies between Mac OS and Linux (since they're writing their own drivers this isn't guaranteed)? Is the scheduling done similarly? Are there any special hardware modes you have to activate with e.g. binary blobs?
There are a bunch of factors that could affect performance even under the same OS (try underclocking your CPU or play around with schedulers). Given the mostly non-existent documentation from Apple I'd strongly suspect that average-case performance will stay worse on the Linux side for a long time.
The currently top-rated top-level HN comment goes into those details. This release significantly improves CPU power management, to the point where it should be similar to MacOS.
Some of this stuff is handled by binary blobs that get installed/upgraded by MacOS, and are running by the time Linux boots.
With the previous release, power per watt and absolute performance were already better than high-end x86 laptops, so if your question is "is this faster and more power efficient than my other Linux laptop?", the answer is probably yes.
If you're asking if it will beat MacOS's perf/watt in all scenarios, the answer will be no for a long time. However, it is probably already beating MacOS in many practical scenarios.
Right. Everything I’ve seen implies they’re both using the hardware the same way (no “limp mode” or such).
Linux tends to be faster because the kernel is just far better optimized than Apple’s appears to be. Massive sums of money (and person hours) have been spent speeding up Linux networking, file systems, scheduling, etc. and it was all sent back to mainline.
Apple’s kernel team can’t possibly compete based on resources alone. They do their best but MS, Google, Amazon, Redhat, and so many others are constantly improving Linux to squeeze out every last drop of performance.
So in many ways Linux is better optimized. I remember one of the developers posting a few weeks ago about just how much faster code compilation was under Linux, because the file system layer is so much better. It was like 6x or something. Sort of an accidental ideal benchmark for stressing that.
But you won’t be disappointed.
E.g. the performance you can get out of the GPU at the moment is a subset of what you can get out of the hardware. Or as another more generic example, until this latest release CPU boost states weren't enabled due to lack of proper cpuidle driver which resulted in regressed single thread performance.
There is nothing inherent about running Linux that will require it be slower, in some cases it will/is even faster, but the lack of everything being fully supported does actually impact performance right now. It has been getting better with time.
Has anyone tried Steam under Linux? It's quite bad under MacOS (2016 casual games stutter, and there is no 32 bit support).
I know rosetta doesn't exist under Linux, but I don't see any options to run steam / proton under rosetta either.
> Has anyone tried Steam under Linux?
Through FEX, yes https://vt.social/@lina/110068264684987710
> I know rosetta doesn't exist under Linux
It does these days! https://developer.apple.com/documentation/virtualization/run...
CPU wise at least its been on par if not better from its first days. At least this is what the various benchmarks at the time showed.
Makes you wonder about how the rest of the system components will compare when they're finished.
Even with the regressed performance it beat osx in a ton of workloads.
Specifically GPU drivers, which can dramatically impact performance. Especially if I am attempting to run any kind of ML workload from Linux. I'm assuming it's basically a non-starter at this point and one is forced to use Mac OS.
Yes. I put that in the capability rather than performance basket though. As in, you can't access the compute shaders yet, rather than: you can and they're slower.
> I just can't justify buying hardware from a company that is so hostile to developers and hackers as nice as it may be.
While I understand what you’re saying, the developers of Asahi have said before that Apple appears to be going out of their way to leave things open for them.
Apple’s boot security is enough they could easily prevent anyone from ever running a non-Apple blessed OS. But Apple made it fully supported to boot directly into a non-Apple OS as your primary OS.
They’re not helping by giving code or documentation, but they’re not putting up roadblocks. Apple seems to be happy with a stance of benevolent indifference.
It’s not official support. But a number of PC vendors don’t give official support either.
A car with one seat seems 100% complete if your use cases only involve you driving it alone. Asahi Linux is absolutely in that kind of spot right now. For some people there is 0 reason to wait, for others it's not even worth booting. If you have fear it's not fully complete enough, I'd say trust those feelings. At least right now.
There have been multiple reports of the last couple years that Apple has been informally internally helping the Asahi Linux team to make it run well on their hardware. Apple cannot come out and officially support another operating system of course, but they are aware of the interest and are helping make it happen, in an unofficial capacity.
>Apple cannot come out and officially support another operating system of course
Apple officially supported running Windows on Macs for many years.
You can still do so with VMs, officially endorsed. But Bootcamp is RIP. Asahi Linux is not a VM, so not a fair comparison.
I wasn't comparing. I was refuting your assertion.
What has changed that prevents Apple from officially endorsing another OS nowadays when it clearly was able to do so in the past?
My assumption would be support.
Boot Camp was officially supported. It gave them an advantage because it helped people move over to a Mac from windows because they could still run some of their software.
If Apple was to help Asahi officially, I think they would be worried that they would be on the hook either for continuing to support it or for damage done to the hardware by other peoples changes to the code.
Basically I think it’s a combination of not wanting to spend the money and being worried about the potential downsides of doing it.
The upside, in terms of additional sales, is likely very small.
> Apple cannot come out and officially support another operating system of course
Why?
Yeah that's nonsense. They already have bootcamp.
That’s incorrect. Bootcamp has not existed for a while now.
Looks like it still exists, they just never ported it to the ARM chips, presumably because nobody wants ARM Windows.
In any case that's really irrelevant to the main point which is that Apple appears to have no objection to other OSes running on their laptops.
Though I wouldn't be surprised if they don't go out of their way to help and make it de facto impossible through secure boot systems.
> I just can't justify buying hardware from a company that is so hostile to developers and hackers as nice as it may be.
As opposed to what company besides those tiny ones? Almost all of them are closed-source only and drivers have been painstakingly reversed engineered over decades.
Yeah, I'm not gonna run out and buy a new Macbook just to wipe it. But these M1 machines are going to be things you just have lying around sometime soon. If my wife upgrades, for example, I know exactly what I'm doing with her M1 Air. :D
Asahi is designed to be dual-booted right now (in fact you have to go off the beaten path to not do that), and it even uses the native Apple boot UI to let you pick your OS
Thanks to the entire Asahi team, your work is truly incredible and so far beyond my pay grade that words fail me. I honestly recently tried and very much struggled to communicate why I was so amazed by your project when talking with friends.
For anyone interested into the GPU side, I can't recommend Linas streams[1] enough.
[1] https://www.youtube.com/@AsahiLina/streams
Sample-rate shading is exceptionally rare (MSAA is rare-ish these days, but I only know of exactly one title that has shipped sample-rate shading), so requiring a basic compiler transform to handle it, especially when they can do so easily because of their tiler architecture, is pretty sane.
The work of the Asahi team is incredible, and so much fun to watch unfold.
I wander if now that you can get a rack mount Mac Pro with Apple Silicon (launched yesterday, the second coming of XServe), running server workloads on them with Asahi Linux becomes a viable route for some people?
FYI it's silicon, not silicone. Silicone is a type of rubber.
The curse of Dyslexia and crap Apple spell check.
This is what they have support for at the moment:
https://github.com/AsahiLinux/docs/wiki/Feature-Support#m2-d...
Maybe not quite possible to do today but if Asahi keeps up the pace they've made so far I bet it will be soon.
It helps that Apple has popular hardware so there's a good bit of people out there interested in developing support for Apple processors.
If you're doing on-the-road visuals and you want a chunky realtime rackmount to render with, $7k for an M2 Ultra might look like a steal. For hosting purposes and parallel workloads though, there are many other options at much better price/performance points in the same market. If you're not leveraging Apple software and you pay for an Apple Silicon server, you're kinda wasting money that could have gone towards an Ampere or Grace instance instead.
"And yes, the team is already working on Vulkan."
This can do more for gaming on the Mac than Apple's efforts combined.
Are you aware that Apple made it clear yesterday they have a full Direct3D 12 translation layer yesterday?
The biggest problem is chicken and the egg. There is no high-end gaming hardware, there are few games, and there isn’t much demand for the kind of games Windows gets.
But without the hardware it’s not clear the games makes sense. And without the gamers it’s not clear porting makes sense. And without the hardware the gamers and the games won’t come.
Even if Apple released a 100% functional and performant Vulcan implementation tomorrow I seriously doubt it would make any real difference to the gaming market on the Mac.
I think there is more to that. As in maybe Apple hw buyers are not that interested in gaming on it?
For most a console make more sense and if someone wants gaming a gaming PC/Laptop usually a better deal.
As far as I know MoltenVK (Vulkan on top of Metal) is available as part of the Vulkan SDK.
DX12 on MacOS was announced a while ago https://www.codeweavers.com/blog/cjsilver/2021/12/22/were-ge... and started to run recently https://www.codeweavers.com/blog/mjohnson/2023/6/1/unleashin...
But that’s part of the problem.
Everyone who is interested in games (like me) gave up on gaming on the Mac a long time ago because of the other issues.
Sorry, but what's the point? Why not just buy a Linux laptop and have everything work out of the box? Why are Linux enthusiasts putting so much effort into supporting hardware from companies that - at best - ignore Linux? This question is also valid for other manufacturers, btw, not just Apple. So much time wasted doing free labor for hardware companies that will just break your stuff with the next hardware iteration.
For me, the Macbook Pro 14 with a 10-core M1 Pro I am typing on has my favorite keyboard and trackpad I've ever used on a laptop (2nd place was the 2013 MBP and this feels the same). It has the best display I've ever set eyes on.
The battery life is the best of any laptop I have ever used by far.
And the performance for number crunching is as high as any x86 machine I had previously and per watt it blows everything out of the water. And it is dead silent while doing so, whereas every x86 'work' powered laptop I ever used would wind up sounding like a jet engine with my workloads.
So for someone who runs linux.. if they want to run it on hardware that is this nice (to me at least)... this is it. This is worth it. It's worth developing for.
Also, Apple is going to stick with the M-series SOC's for a long time now that they have switched. And they tend to keep hardware interfaces for a long time too. So the development of Asahi now will bear fruit for... at least the next decade I'd say.
I still use OSX for daily activities, but the kernel Asahi is developing may be my plan to stretch this 2021 M1 MBP 14 out hopefully to the year 2030, as MacOS moves on. My 2013 intel i7 Macbook Pro made it to 2021... 8 years of daily use and world travel. I was beyond the moon with that product performance and I'm expecting similar from these new macbooks based on my current 1.5 years of use.
> Why not just buy a Linux laptop and have everything work out of the box? Why are Linux enthusiasts putting so much effort into supporting hardware from companies that - at best - ignore Linux?
Almost all laptops sold with Linux pre-installed or with support advertized only work well with Linux due to volunteer work similar to what is being done with Apple's stuff right now. Almost everything is proprietary with close to zero upstream support, you just don't notice it because the work has already been done.
In an OSS ecosystem, manufacturers inherently have to use work that has been done by volunteers! The real question is whether they invest their own resources too. For example, System76 has created their own distro, Pop OS. My team and I were happy with the laptops purchased from them and the OS integration provided a smooth experience.
Are there commercially available arm-based linux laptops? I'm not after horsepower just decent battery life and enough computing power to do non-AI programming.
The point? There are no other ARM laptops that are even close to being competitive right now. I want a good fanless laptop. My only choice is Apple. Asahi made it possible and I now have a great experience. What's the problem?
> Why not just buy a Linux laptop
Please tell me where is it because I will buy it instantly, and I’m only half kidding.
I don’t see how making linux available on possibly the currently available best laptop hardware any different to the previous decades of hacking a working wifi driver into the kernel. It was always an uphill battle, and we should be thankful for those who take up the hard work!
> Please tell me where is it because I will buy it instantly, and I’m only half kidding.
System76? Star Labs? Purism? HP Dev One?
System76 doesn't ship reasonable screens (they are all 144Hz 1080p):
https://system76.com/laptops-ultraportables https://system76.com/laptops-powerful
The only exception is the Bonobo, but it comes with a discrete GPU, so its battery life + weight are going to suck. Also, its keyboard is off center.
Most of the star labs machines have low resolution displays, but I can find nothing wrong with this one. If you choose AMD, a reasonable config is $2600, which is comparable to Apple. However, they are only building 10,000 units, and taking preorders, so it might be unobtainable:
https://us.starlabs.systems/pages/starfighter
The purism offering seems OK except that it is a 10th generation intel, and those were extremely bad, even by recent intel standards. Maybe they'll get an AMD refresh out the door with the same ergonomics.
The HP has an off-centered keyboard and trackpad and a 1080p display.
So, of those four vendors, there's one model that's vaguely competitive, but it's a limited production run pre-order.
HP Dev one works well, kudos to System76 for the desktop env.
It has the worst LCD panel I've seen in many years. It's not the resolution just poor contrast and poor color accuracy. It pains me to see it.
They went on sale recently, I bought one, and shortly afterwards (a few months ago) they stopped selling them.
System76? Framework? Even Dell has a number of Linux laptops.
If your requirements are "must run an Apple designed ARM processor", then yes, your choices are pretty limited.
ANY arm processor, do tell
Any? https://wiki.postmarketos.org/wiki/Lenovo_IdeaPad_Flex_3_Chr... is dirt cheap and runs PMOS quite well. Actually Chromebooks in general are a good bet when combined with PMOS - or even not combined with PMOS, if you can work with a normal Linux VM on top of ChromeOS.
well obviously not a machine with 4GB of RAM running outdated CPU
You said "ANY arm processor". If you'd like to further constrain the exact requirements, that would be a useful clarification. (Although, I suspect that some Chromebook will still work)
It's to respond to the remark that people are simping for Apple while in fact, they're just after a good arm-based laptop.
Thinkpad x13s is an okay option, but Qualcomm is IMO more cursed than Apple
What do you mean?
system 76 makes linux ready hardware. personally I'm waiting for a framework which does have some linux support
Apple makes good hardware and makes good engineering decisions. Sure other laptop makers have low end and high end options. But frustratingly often match apple on 50-75% of the features. Sure some have nice still aluminum bodies. Some have screens that match or beat apple's. Some have nice centered quality touchpads and keyboards. Some have great battery life. Some have great performance.
Very few match on all aspects. I'd tried a few and always had one terrible issue. Terrible battery life, lousy screen, and/or terrible touchpad. Apple does seem willing to make improvements without as much worry about backward compatibility. There are a few that match on everything I care about, but often cost more than the Apple.
People like to complain and mention byzantine purchase methods. Wait for a lenovo sale, buy the bare bones model, apply the discount code, then buy dimms and SSDs from random bargain basement sellers. Oh and buy the linux compatible wifi card and do surgery on your laptop to get wifi working after suspend.
MBA is pretty compelling mix of performance, size, cost, and battery life. Unlike any x86-64 laptop, you can pay $500 more or so and get double the memory bandwidth. Or another $500+ and double it again. Definitely makes the macs better any PC at some workloads. Sure some x86-64 with a nice discrete GPU is way faster ... when plugged in to wall power.
They don't exist, or if they do, they rely on far less (motivated, whatever that motivation might be) community folks to bring-up support. Which often happens are the device is a generation or two old.
Non-sbc aarch64 devices with good upstream support are damn near non-existent. I understand the Linux user market is small, but it's baffling to me that companies can't invest a small amount in having coordinated upstreaming efforts.
That said, I'm so delighted with my G14 2022 experience that I might just buy the 2024 version of it's all AMD again. I have better hardware control in Linux than Windows, I can play Halo Infinite in Linux (praise be Proton) and Windows. But... I suspect I'm going to be able to do the same thing on a likely more-battery efficient and nicer hardware M2 MBA by the end of the year. Nice to have options.
Edit: it's really soulcrushing to think about. Sdm845 devices are insanely cheap and yet upstream support still barely limps along, mostly with volunteer/donated efforts.
What Linux laptop has comparable hardware?
> Why not just buy a Linux laptop and have everything work out of the box?
Because these things are rare as gold in the first place as being Linux ready isn't a focus for most OEMs, sometimes severely lag behind the competition in feature support (e.g. limited to UVC webcams with crap quality), have serious availability issues (Framework), you have a tough time getting service or spare parts, or barely any resale value, or limited choices in screens (which is the one and only thing keeping me from a Framework - who the fuck wants a 3:2 screen?). Also, tough luck getting firmware updates for embedded components.
Apple devices, spare parts and repair centers, in contrast, are widely available across the world (okay, maybe not in places sanctioned by the US), firmware updates come around when needed and hold their resale value for years.
Are there any PC laptops, running Linux or Windows, that you can buy for a similar price that run as long, as cool (temp), as fast, and with a screen as good as an M2 Air?
I know there are faster machines, but they’re usually hotter. There are cooler machines, but they’re usually not as fast. There are cheaper machines but the screens are worse. And there are long battery life machines but they’re usually low performance Chromebooks.
I don’t know of any other machine that made the trade-offs to end up offering a similar combination. At any price.
And I said nothing about noise or thinness.
Once you’ve used one, it’s a very seductive combination of features. I don’t want to go back to “6-8 hours of battery life is enough”, or hearing fans, or using my laptop as a hand-warmer in summer. And I don’t understand how 1080p is still so common.
Far superior hardware.
With that attitude, Linux and its ecosystem wouldn’t exist. But to answer your question; because they can and are having fun doing so.
The point is to have Linux support on some of the best hardware available at the moment.
I can’t wait for M2 (pro) support for my MacBook Pro. I was long term Thinkpad/Arch Linux user and really want to go back to such setup. Sadly I didn’t find anything better hardware-wise than the MacBook but I love Linux.
I know they are focused on getting it to a good quality on M1 first but eagerly consuming all project updates! Good job team!
I'm still waiting for Apple Silicon that's fast enough to run Gentoo Portage at Arch Pacman speed.
Kinda irrelevant but Asahi is the topic on HN that gets me excited the most. Can't wait to have such a great hardware for my daily driver.
-- a happy ThinkPad Debian user
When the Asahi alpha released dropped last March, it was stable and functional enough to become my daily driver. And since then the pace of development has been steady and impressive - I actually look forward to every 'pacman -Syu'.
If someone told me 20 years ago that I'd be using an Apple-made ARM workstation running Linux to do all of my development work today, I would have never believed them.
Super curious if it is possible to run nvidia gpu on new Mac Pro using Asahi Linux? Can it be vritualized with pass-through?
The one announced yesterday?
I think the main problem would be where would you get ARM drivers?
On Intel people were able to existing driver blobs. But if there are no official ARM blobs anywhere, it would be a much bigger effort.
Is the idea games? Seems like there’s already some interesting Wine and CrossOver developments.
No, I need nvidia for ML
Is the desktop GPU-rendered by this point? What's battery life like (for those using Asahi daily)?
Thinking about taking the dive...
It is GPU accelerated and very smooth since late last year. Battery life is about 6-8 hours on an M1 air for me, and should improve significantly with this latest upgrade (haven't measured yet).
The installation experience is very smooth, so it's very easy to try it out. You can nuke it afterwards if it doesn't work for you.
It appears so.
From last December:
> "This release features work-in-progress OpenGL 2.1 and OpenGL ES 2.0 support for all current Apple M-series systems. That’s enough for hardware acceleration with desktop environments, like GNOME and KDE."
https://asahilinux.org/2022/12/gpu-drivers-now-in-asahi-linu...
When Vulkan drivers are ready maybe the Asahi Linux perhaps also be ready to run some SteamVR apps https://github.com/ValveSoftware/SteamVR-for-Linux/blob/mast...
I'm surprised that Vulkan wasn't the target, which could then have something like Zink layered on top for OpenGL.
This gets mentioned basically every time. The gist is that:
- OpenGL is a much easier target to support, in terms of having a functional desktop. Remember, hardware-accelerated apps have been runnable on Asahi for months now, while Vulkan support is still a while away. - Most of the work being done is common to OpenGL and Vulkan, so it's not exactly a ton of wasted effort.
Zink requires a fairly advanced Vulkan driver, with many complex optional extensions. A basic Vulkan driver wouldn't cut it. So getting basic OpenGL working is much less work, especially since they can share a lot of work with Mesa's existing OpenGL drivers.
Their work tempts me to get a Mac some day just because I know I can run Linux on it.
Recently, I've been thinking about using Asahi as the host system running on my M1 MBA and run everything macOS in a vm. Does anyone have experience with that? How stupid would that be?
People have gotten QEMU with KVM running VMs and other people have gotten macOS ARM64 running under QEMU... so I think technically possible, though I haven't actually tried going that far with it. I don't think you'd get virtual GPU acceleration support working in the current state, so performance would be pretty god awful.
Does anyone know how efforts are going to to get Asahi's kernel changes merged back upstream? Has anything been merged upstream already? Is there a roadmap for getting things upstreamed? Are they pursuing an incremental approach to upstreaming individual components or do they have to prove that everything is flawless on Apple silicon before anything can be merged?
The feature table here https://github.com/AsahiLinux/docs/wiki/Feature-Support also tells you where along the path something is in terms of being stable and making it into the kernel. If it just has a release number like "5.19" then that means it was successfully upstreamed as part of that kernel version. Generally, the Asahi team tries to upstream as they go, but if they are uncertain about the overall architecture of a driver they'll hold off to prevent ossifying certain design decisions too early. An example of this is the GPU driver.
This information is on their website wiki. See https://github.com/AsahiLinux/docs/wiki/Feature-Support
How's proton support nowadays? The new M2 15-inch macbook Air looks really appealing
Proton is a wrapper/tweaks to wine, right? Thus x86-64 windows games can run on x86-64 linux. Thus the steam deck and many happy linux gamers. Apparently it's good enough to run a large majority of steam games, I believe I remember something like 90+% of the top 50 games on steam.
I don't think it helps at all with running x86-64 code on arm.
For running x86-64 on arm you have FEX. You can combine FEX with wine to run standard steam games on Asahi linux.
Wow, excellent. I'm trying to justify replacing a 2015 desktop with either a mac studio or a ryzen/zen4 desktop. I'm not a gamer, but do occasionally fire up steam to play something old like Xcom 2 or Majesty.
Knowing that at least some x86-64 steam games could work is promising, thanks.
For me it all comes down to the extra memory bandwidth.
A recent metal port of llama is pretty tempting and being able to run GPU accelerated LLMs with greater than 16GB (mid range GPUs) or 24GB (highend/RTX 4090) on the mac studio is interesting. $3,600 (for 96GB ram) interesting, not so sure.
Check out this stream: https://www.youtube.com/watch?v=CJSfFzsU75g
Where many steam games where tried live and debugged to run FEX + wine + Asahi GPU driver through its paces. Highlights include Portal working, Fallout NV working. Crysis running with graphical artifacts and lots of other games. The main limitation currently is simply GPU driver feature support. To run modern D3D12 games you basically need Vulkan which Asahi GPU driver doesn't support yet.
Wow, 12 hour video, I do wish a small summary was posted. Even just a sentence or two, like your post, thanks.
Sounds like the OpenGL work can mostly be reused for improving the Vulkan driver and sounds quite promising.
It can run under FEX or box64 though
No Vulkan drivers. Will need that before anyone tries to do much with proton.
How much do you need macos around to use it?
Sounds like the original install and then for firmware updates? I’d like to keep telemetry to a minimum and not use macos if possible.
Yes, that's right; I generally pop back in to macOS every so often to check for OS updates (which update the firmware as well - no way to just update the firmware).
I seem to remember there being a tiny recovery partition or cut-down version of macos or something maybe from boot CD (we have an old iMac).
Also found this:
https://apple.stackexchange.com/questions/311947/how-to-upda...
Still I'd rather have no MacOS.
There is a recovery partition (in fact, it's used when installing Asahi) but I don't think it's possible to update the firmware from there. Boot CD might work. You could try asking on r/AsahiLinux.
Awesome work. Now if we could just get those parts in a modern, modular laptop with replaceable components. Frame.work has raised the table-stakes for laptops.
Related: does anyone do development on a Mac in a Linux VM? If my dockers are already running in a VM, why not go to the next logical step?
My work issued laptop is a Macbook Pro M1.
I typically work on my work issued linux workstation (pretty old now and less powerful than my Macbook Pro), but often need to do development from my laptop too. I run a Fedora arm VM in UTM full screen. It generally works well, although I'd much prefer to have native linux on my laptop. It would be nice of mac os didn't interrupt full screen mode randomly and allowed UTM to capture all keys and touchpad gestures.
I personally am incredibly unproductive on a mac, and have no idea how anyone does anything with how terrible the window management and virtual desktops are. Plus, I find all my linux based tools that I am accustomed to just work so much better under linux. So for me, even a VM is still a huge leap in productivity.
Utm is there if you want reasonable size desktop without fighting with qemu settings / setup. https://mac.getutm.app/
I tried utm early on with an m1 but had many graphics issues that parallels made work out of the box.
I need to try it again, bet it works out of box now.
I have done so for years on my 2013 x86 MBP, using VMware Fusion.
It's really nice being able to four-finger swipe between Macos and Linux full-screen desktops when I want that.
But usually I SSH into the Linux VM so that I can use iTerm2 as my terminal, and use Mac native GUI Emacs (railwaycat & Mitsuharu Yamamoto's version) to have its graphical features working alongside my iTerm2 and Firefox.
Linux in a VM runs some things much faster than the host MacOS on the same laptop! In particular filesystem intensive builds. I found some Docker-based test system I had to use for work ran an order of magniitude faster in the Linux VM than with Docker on the Mac host.
I did test running Linux natively on this MBP (no MacOS host) but found to my surprise the battery drained faster than using it in a VM, and the VM is almost as fast. So I stick with the VM and I get to use both OSes side by side.
I don't use the VMware file sharing as it is extremely buggy and regularly corrupts files when operating on them with Git or some other activities due to buggy cache settings which cannot be set to non-buggy values.
Instead I use some crazy combination of Samba and NFS which works reasonably to get the right permissions and ownership as seen by each side of the other's files. The upshot is two-way file sharing works, but this was actually difficult to make reliable.
(I found the Mac SMB2 client is prone to occasionally deleting files that aren't even being accessed by it at all, but are in a directory where other files are being updated. A shocking bad SMB2 protocol bug in the Mac client where it sends the wrong filename for renames occasionally, using a name from a recent directory listing. It took me a year to diagnose that and I thought it must be a bug in my code in a critical Linux database application I was writing, which I didn't dare deploy upstream for all that time, because important random files like invoices and data files would occasionally disappear and be noticed days later causing me to doubt myself. That is until I did a very deep dive with every debugging tool you can imagine, to find the true cause was MacOS when I edited unrelated files, and my application had been fine all that time to deploy to prod. Solution was force MacOS to use CIFS instead of SMB2.)
I hope to replicate something like that when I get an ARM Mac at some point, though I hope the MacOS SMB bugs are fixed. Currently holding out for an M3.
I'm using a multipass docker setup. It is faster than my previous setup, but when running make -j from inside the container top says that only 1 of 8 CPUs is getting scheduled to userspace at a time.
Also, the bind mount of the external MacOS directory is extremely slow. I do out of tree builds so that the builds land in ext4.
I haven't gone the next logical step because (1) it is plenty fast for rust development, and battery life is fine, and (2) I'd like webcam and speaker support.
Also, I just checked, and I have a 2023 model, and the installer status is "WIP", so I guess I'll be waiting a bit longer.
I use Parallels w/ an Xubuntu VM running some Docker projects and am pleased with the results.
Filesystem performance used to be better than Docker Desktop, but they seem about on par w/ one another now.
> Filesystem performance used to be better than Docker Desktop, but they seem about on par w/ one another now.
They made an update that claims performance is comparable.
I tested realistic workloads last week and docker desktop was still meaningfully worse.
If your work computer has antivirus, the performance of docker desktop will be even worse if exceptions aren't or can't be added.
Filesystem/disk performance? Although that's probably an argument against running Docker on arm64 Macs too...
Docker desktop has arm64 builds.
I do all of my development work on my x86-Mac in a VMWare VM. Works very well. I am using Fedora.
I used to do this before switching to a Linux laptop. It worked mostly fine.
Can Asahi Linux run on the new Mac Pro? Would I be able to use the GPU to run LLM models?
They’ve never done an M2 Ultra, as no one (anywhere) has one yet.
I believe they said they will work on it but they can’t promise a support date because they don’t know how long it will take once they have the hardware. And I don’t know when they’ll get the hardware.
So basically the recommendation was “don’t buy it to run Asahi yet, you’ll probably be waiting a bit”.
Is anybody daily-driving any M2 macbook pro on Asahi Linux? What is your experience like?
It wasn't until this morning installing Asahi on an M2 Pro was supported in expert only install. Prior to that it hasn't been supported at all. There is still some work to go on the Pro/Max/Ultra SoC family.
Thank you Alyssa Rosenzweig.
I heard the Asahi Linux code is super dry.