Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

What I’m interested in, which may probably take years (or never happen well), is running Linux natively on these Macs after there’s no more macOS support from Apple. Typically Macs stop getting software support at around 7-9 years after first release. Hardware does tend to last longer and can serve some purposes beyond that.


Me too. Interestingly 2 MacBooks from the Intel era were amongst the best machines ever, objectively, for Linux.

The MacBook 2,1 is almost unique in being supported by Coreboot and Libreboot. Furthermore, because it had an Atheros wireless card, you didn't need any blobs at all. Only two old Thinkpads are comparable, among x86_64 machines.

The MacBook Air 11 Late 2012, which was used as a daily runner by Linus for a number of years, was a pure Intel machine. Except for a weak Broadcom card, it was flawless with a stock kernel. Plus, it was silent, small and fast. The only comparable machine in terms of silent operation, cost and Linux support was IMHO the Xiaomi Notebook 12.5, which was released quite recently.

The problem with ARM Macs is not just secure boot. The secure enclave chip already gave serious trouble when trying to run Linux on their last Intel machines, as e.g. the keyboard doesn't show up as a standard USB device. So I don't have high hopes.


Completely open is cool, but it’s a high bar to pass. A lower bar might be, can I just install Ubuntu and use it as a daily driver? At least up through the mid-2015 this is the case.


Oh yes, I'm happy with that. The problem is that might be even hard to achieve with the T2 chip.

It's sad because the new MacBook Air seems superior to all competitors and the price is quite fair after academic discount.

Also, Apple makes it easy to buy any keyboard distribution from any location, which is really cool. With many brands, I need to hack the system and shop from NL to get a US ANSI keyboard here in EU.


Standard layout for dutch macs is ISO. Although English-US with ANSI is available


I meant for other brands where I have to get them from NL to obtain a US-like keyboard. Otherwise, they ship me the local EU keyboard.


Yeah, in particular the Mac mini has a lot of curb appeal as a 10+ year micro server. Considering it's plugged in and the fan should only spin up occasionally, it should last a long long time.

I'm pretty sure we'll have Linux running on it by then. Perhaps just CLI only.


My 2011 Mac Mini is running under the TV, serving files, running some small docker containers. It works fine for that workload.

I just use VNC into it to manage anything. Another little machine that does what it needs. It's quiet and doesn't need much.


Why not just use an Intel NUC?


Last I checked, a well configured NUC was nearly as expensive as a Mac mini and almost certainly slower than this. Depends largely on what you are looking for though. From the sounds of it the NUC also has a loud fan? (For some reason I thought they were fanless).


Yeah, both the Mac Mini and NUC have fans. The Intel NUC will definitely be slower and hotter (thus louder), but I still think it's a better choice if you specifically want to run Linux on it.


>What I’m interested in, which may probably take years (or never happen well)

Probably wont take that long. Apple gives instructions to disable strict boot check and boot from any OS. The rest are the drivers. For graphics, they can start just using a framebuffer...


And 7-9 years is a fantastic support lifetime and yes, I know some folks who still use their 9 year old macbooks such as grandparents etc without major issue.

You have new android phones that ship with software 1.5 years old and NEVER get updated.


7 to 9 years is fantastic, but not really that out of the ordinary for this industry as far as I can tell?

My 2011 model Lenovo came from the factory with a 2012 copy of Windows 8.

Lenovo released BIOS and driver updates to the 2011 machine at the end of 2019 and continue to do so. Windows 8 is EOL in 2023 with free upgrade paths to 8.1 which has a further out EOL date, meaning that the OS is still supported by both companies, as well.

Admittedly, however, your Android example is dead-on. Android seems to have always struggled with software updates; I blame version fragmentation along with very cheap (free in most cases) licensing costs, which spur on lower-quality vendors to take on the product.


I agree but I'll tell you (as an apple fan in the ecosystem) that sometimes this kind of "let's abandon the old things" has some drawbacks. For example, the changes in the https behaviour and support for the websites, made an old mac useless because safari didn't support the new TLS versions and there weren't any decent alternative for an old system like that (we're talking about a 2004 mac if I recall correctly, intel 32bit, so yeah, I know it's old in many ways).

The only solutions was to install linux and it worked.

So, even though I'm fully committed to the ecosystem and I love each and every apple product I still do think that having to trash a machine after 10 years isn't so good considering that those machines are usually still capable to work daily.


I'm typing this on a 2011 MacBook Air. I don't use this as my daily driver, but I use for lightweight day to day things.


These M1 Macs are more like iPads than Intel Macs. Jail breaking them would be the first step, but after that one would need a ton of reverse engineering to write all the drivers for all the little custom HW built into M1 SoC.

I have never heard of an iPad running Linux (and iPads been around for a long time), so chances of M1 running Linux natively is slim.


What would native support provide that virtualization does not?


Performance, a teensie bit of insulation from long-dead software rot, and the warm fuzzies you can only get from running a free software stack.


I’m guessing there won’t be a significant performance advantage, not insulation from rot.

The likelyhood that Apple’s drivers at EOL outclass the driver support the community can provide seems high.

On the other hand, I agree with the last point. There is an aesthetic dimension that can’t be ignored.


The community does a pretty decent job of supporting other Apple hardware; often vastly outperforming default MacOS (or Windows in many tests). Graphics is a particularly soft target, eg: https://www.phoronix.com/scan.php?page=article&item=macos101...

I know —and apologies for the originally unintended, but retrospectively way-overused pun— I've compared an Apple to an orange there, and I realise it's very different hardware... But it's also not. It mostly depends on access. If developers can shim in a loader, and it doesn't require soldering, it'll be popular and get attention. These chips look too great to ignore.

Not my downvote, btw. Hopefully they'll toss an opinion in too.


That’s very good to know - I have an iMac 5K 27” that is only a year or so away from losing support from Apple, which will become a Linux machine at that point.


Virtualization doesn't always play nice if you're doing performance intense/measurement work, although I wouldn't even try doing that on Apple hardware to start with


I don't know why you believe it can never happen well, as linux already run in ARM processors for some time now.

Obviously it would take another canonical to do grunt driver creating work. But I suppose that if we don't see a quick response from intel or AMD, this hardware will take off and people will write the drivers.


It's not just a question of driver support. Apple is doing a lot of custom stuff with their architecture that might not have an analogue on current ARM systems, meaning that it might require not just new or updated drivers, but effectively a completely new architecture variant added to Linux before you can even boot.

Also, with the T2 chip in Intel Macs, there was a lot of nonstandard behavior[1], like the keyboard not being just a standard USB device.

Linux can run on ARM processors, but it's the architecture that can make a huge difference. Heck, even Apple's Lightning-to-HDMI adapter has an ARM chip in it[2], but that doesn't mean you can boot it to Linux very easily.

[1] https://news.ycombinator.com/item?id=25221804

[2] https://www.theverge.com/2013/3/1/4055758/why-does-apples-li...


> but effectively a completely new architecture variant added to Linux before you can even boot.

This sounds wrong. At most it may need a custom bootloader, which can then load a regular arm64 kernel plus the required drivers for Apple specific peripherals.


The laptop works without updates.


It might keep working for a while, but Macs depend on external services to function properly. What if Apple replaces OCSP in a future release of OS11 and turns off their existing OCSP service?


Is the OS secure abandoned without 2-3 years of updates to any new issues discovered?


You expect a consumer device to be support for more than 9 years? That’s not practical.


Older macs could run Linux pretty well. I even had Gentoo running on an ancient PPC mac yonks ago. A few years back, I got Linux working (sorta) on a MacBook 14,3

https://battlepenguin.com/tech/linux-on-a-macbook-pro-14-3/

With all the security on the newer Apple chips, I wonder if we'll ever see Linux boot on these things natively, much less get actual hardware driver support.


Apple started putting the T2 chip in the Macs 2 years ago. I suspect they plan on locking down the platform soon.


I don't really get this narrative about the T2. Apple removed the T2/integrated it into the M1 and didn't lock down things any further. If they were going to lock it down, this would have been the time for it, given there's no Bootcamp for Apple Silicon. Yet Apple specifically allows you to sign custom kernels offline on Apple Silicon Macs.

People also seem to be convinced that the T2 was a "locking down" measure, while all that's really missing is a Linux driver to interface with the T2. Surely Apple doesn't care at all about actively supporting Linux (and publishing specs), but they also don't lock this down.

The only support I see for the idea that Apple will "lock down Macs" is that for some reason they're irrationally hell bent on turning Macs into iPads. I don't see how that idea holds up if you inspect it closely though.


I'm not trying to peddle a conspiracy theory, but the slow progression "boil frogs" method is how this sort of thing is usually done. I mention the T2 because secure storage for secure boot is a pretty obvious way to implement.

The percent of revenue coming from Macs continues to go down. I'm just assuming that at some point, the only (financial) reason the Mac exists is for Xcode.


Apple's Mac category revenue was $9 billion in the latest quarter, up 29% year-on-year—growing faster than any other category. And this was before the M1 launch.

At this revenue run rate, a standalone Mac company would be in the Fortune 100, next to companies like Nike and Coca-Cola.


That's a quarter. See:

https://cdn.statcdn.com/Infographic/images/normal/8817.jpeg

https://images.macrumors.com/t/hIKaZ3dlMYeHuw-3yqiPbLiHtnc=/...

And, yes, it's still significant, but it's also lower margin revenue than other Apple products.


Pretty sure Mac is going to do well this quarter, too. And surely Apple Silicon means that margins are going up.


It’s a conspiracy theory.

Any attempt to improve security is of course going to follow a progression. All software development is progressive.

The leap from there to a ‘boiling frog’ i.e. that this is being done so the user won’t notice a hidden agenda, is a conspiracy theory.


I suppose, if you consider any speculation to be a conspiracy theory.


It’s not ‘any speculation’.

It is a theory that a group of people with power have a hidden agenda that is contrary to what they have public stated, despite the facts being consistent with the public claims.

That is pretty much the definition a conspiracy theory.


A group of people that already understand and take great advantage of locked down platforms and services. And we're talking about a product (Mac) that was specifically much more locked down than its predecessor when it was launched.


A group of people who have said that they see the Mac differently from iOS, and who have explicitly said they don’t plan to lock it down.

For you to be right, this must be a planned deception.

That is a conspiracy theory.

It seems like you actually know this and are just trying to argue that that you’re right that there is a conspiracy.

For what it’s worth - I don’t think there is anything fundamentally wrong with that. Conspiracies are sometimes real.

I’d just rather be open about what we are saying.


>For you to be right, this must be a planned deception.

Sure, you're right, it's conspiracy theory -- but one doesn't have to be a conspiracy theorist to predict further anti-consumer action from companies that have a long and sordid past with anti-consumer actions.

Apple , quite literally, wrote the playbook for establishing gilded cages and vendor lock-in within the (modern) computer industry.

I mean.. most companies in the world don't generate enough ire that the public get together and maintain a wikipedia page about the criticisms that apply to the company separately from the parent page.[0]

But absolutely, I agree -- it's conspiracy theory. The problem I have personally with conspiracy theory regarding Apple is that over the past 30+ years i've seen a lot of it turn into fact gradually over time.

I have very few doubts that the trend will continue onward. -- as such I have little problem with entertaining conspiracy regarding Apple, as long as it's not absolutely insane.

Apple seems to be masterful at the magicians' game of misdirection. That's the best way I can put it.

'Ignore the man behind the curtain.'.

The most modern example?

Let's talk about how fast and cool and battery conserving their new processors are without taking into account the vast amounts of software breakage that came along with the architecture change and the future software landscape reform that's going to take place that will establish Apple customers even more firmly into territory they can't ever hope to leave.

...and this is coming from someone who , historically, has loved ARM. MY problem with it is that the architecture change, IMO of course, is going to undoubtedly be used to leverage and strong-arm customers into Apple-centric 'app-store' interfaces, pulling the market apart just like in the PPC days between Apple OK-d software and the rest of humanity -- for the sake of a lot of quick bucks for Apple and a lot of additional developer/customer friction for the rest of the world.

We'll see, I guess. I like the new M1 on paper, I just trust Apple about as far as I can throw them.

[0]: https://en.wikipedia.org/wiki/Criticism_of_Apple_Inc.


The presence of a Wikipedia page critiquing the worlds most valuable corporation doesn’t seem like evidence of anything much.

It’s worth noting that many of the criticisms on that page, while valid, are true of all their competitors too, and of comparable capitalist businesses.

What that page leaves out is that in many areas, Apple has made more progress at mitigating the problems than anyone else.

I have spoken to activists about this discrepancy, and received the answer that they are aware of this, but it is politically more effective to critique Apple because they are the market leader, than to critique their competitors.

I don’t really agree with this tactic, but I understand it.

Nevertheless why even bring this page up?

It seems like you think it supports a generalized narrative of ‘Apple as bad guy’, which then makes the conspiracy theory more plausible.

This is unsurprising because it is the normal way conspiracy theories are supported: some real problems are tied together with woven together with exaggerated claims in a grand narrative with a villain at the center.

We are asked to believe in the conspiracy because some of the things that are part of the theory are based in fact, despite the fact that the rest is just innuendo and naked assertions.

That is what you are asking us to do here.

See examples:

[1] “Apple seems to be masterful at the magicians' game of misdirection. That's the best way I can put it.”

Take this for example. Here you just affirm the consequent. Of course, as I said earlier for you to be right there must be a deception, and here you say that Apple are masters of deception.

But really that’s a naked assertion. If there is a conspiracy this would be true, but there is actually no evidence of it.

[2] “Apple, quite literally, wrote the playbook for establishing gilded cages and vendor lock-in within the (modern) computer industry.”

Obviously this is fantasy. If Apple had literally written such a playbook, you would be able to present it as evidence.

What they have written and presented literally, is a large body of work on how to produce a secure computing environment that mitigates most forms of security threat and is therefore trusted by consumers as a platform where they can purchase software.

That is literally what they have written.

Again you are just asking us to accept a conspiracy as if it were fact.


>A group of people that already understand and take great advantage of locked down platforms and services.

They never pretended otherwise. It's a commercial company, with its own platform.

It's not a generic OEM manufacturer for generic devices to install various OSes on.

Their whole pride is their software/hardware integration.


Sure. We're pretty far down the rabbit hole, but I'm responding to people that think my suggestion that Apple might lock down Macs is unfounded. You seem to agree with me that there's some reason to believe it's plausible.


For what it's worth, these are all the same arguments people have made regarding the TPM chip over the last ~15 years:

https://en.wikipedia.org/wiki/Trusted_Platform_Module

It also reminds me a lot of the Palladium controversy of ages gone by:

https://www.zdnet.com/article/microsofts-palladium-what-the-...


Isn't the TPM a large part of why it's difficult to run anything other than iOS on an iPad/iPhone?


iPhones don't really contain any hardware that does not have more or less a conceptual equivalent in other smartphones. I suppose with "TPM" you mean the "Secure Element", which on iPhones and modern Android phones contains the disk encryption keys.

Like a lot of other smartphones, the main difficulty in running something else on an iDevice is the locked bootloader. It won't run anything not signed by Apple. If an exploit like checkra1n is used to defeat this, iPhones can run Linux in principle [1], of course practically restricted by a lack of a large amount of drivers required to run Linux well.

This iDevice OS lockdown completely reduces to 1) the locked bootloader which must be "defeated" - which M1 Macs contain a built in tool for - and 2) the only OS with drivers for the hardware being Apple's Darwin. So no, the TPM isn't a large part of why it's so difficult to run anything else on an iDevice.

[1]: https://blog.project-insanity.org/2020/04/16/running-postmar...


You probably mean the Secure Enclave. The Secure Element contains payment applications and keys for Apple Pay.


No it was an intel thing for secure booting windows in x86 PCs back in the days when Microsoft was top of the world and everyone thought Wintel were evil monopolists trying to lock Linux out of the PC ecosystem. I think there are still TPM like capabilities in intel chips now, but they rebranded it?


I had assumed TPM was generic enough to refer to the T2 as well.


Delayed reply, but no, TPM is a specific international standard:

https://en.wikipedia.org/wiki/Trusted_Platform_Module

iPhones, iPads, etc. do use a similar concept though.


>I'm not trying to peddle a conspiracy theory, but the slow progression "boil frogs" method

Yes, but another name for the "they're using the boil frogs method" is "slippery slope fallacy".


How's that? Slippery slope depends on a "relatively small first step". Apple already has other hardware and software platforms that aren't open.


The boiled frog fable [1] is false.

[1] https://archive.vn/mu6T


As are many illustrative, and still useful, metaphors.


In other news, the cow did not jump over the moon


With their M1 based devices, they now explicitly added support for you to authorize alternative kernels to boot from (on your local machine).


Is this better than the previous T2 Intel Macs for Linux support?


Previously you weren't able to do that, but you were able to disable booting without signature checks. I'm mainly basing this info off this sole tweet: https://mobile.twitter.com/never_released/status/13263157410...




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: