People should get on there and support him financially like I have.
To maintain freedom takes work, effort and constant vigilance.
The least we can do is contribute to the work and effort.
Japan-based developer Hector Martin has formally kicked off the crowdfunded Asahi project, which aims to get Linux running on Arm-based Apple Silicon Macs. "The goal is to bring Linux support on Apple Silicon Macs to the point where it is not merely a tech demo, but is actually an OS you would want to use on a daily driver …
The term is from the USA, from cars, i.e, a daily driver might be a reliable Toyota you use to commute and shop, as opposed to the vintage muscle car that you keep in the garage because you're doing it up as a hobby.
The term is also borrowed by wearers of wristwatches - a 'daily driver' might be a Casio, as opposed to your grandfather's watch that you only wear on formal occasions (for fear if losing it damaging it).
Freedom to do whatever you want (1) is not the same as freedom from others telling you what you can do (2). One can have an as free as it can possibly be (type 2) computer now. (Do you fab you own components? no, I thought not). Companies exist that let you do whatever you want with their hardware. Do you run openbios? It all starts there UEFI or legacy boot.
Apple do make nice hardware, so I'm told, but if you want those particular shiny toys you are asking for type 1 freedom. There's no evidence that Apple want you in their orchard. They might let you carry on forever but if you threaten their business model I don't think you should be surprised if they do something about it. Here the past is definitely a guide to the future.
I would rather continue to fund (not breathtaking amounts) and support type 2 freedoms but good luck ( no implied "with that one")
I agree with a lot of what you say.
Personally, I won't have Apple hardware in the house. I hate them as a company and their relentless efforts to stymie repair. Their kind of control and disdain for their users is hardly unique to Apple though.
However, this is one valiant effort to bring freedom where little enough exists and I applaud it.
Whilst I agree Apple's customer service has gone seriously downhill, I simply can't abide Windows 10, having used it for the past 3-4 years. The constant forced updates rendered my laptop so slow plus the the hardware incompatibility that comes with having to run on so many different components was a pain too. Yes if my new Mac Mini ever has a hardware failure, I'm sure after the warranty runs out, they'll do their best to fob me off or try to charge me £400 for a simple repair but at least, in the walled garden, it's a very stable system.
Did you read the article? Apple not assisting in supporting other software isn't a block. Indeed Apple isn't actively blocking others from running their own software. Quoting the article where the author of the software that is the subject of the author (pay attention in particular to the 3rd paragraph):
Martin said that "Apple allows booting unsigned/custom kernels on Apple Silicon Macs without a jailbreak," which he takes as evidence that "Apple does not intend to lock down what OS you can use."
Apple does control the boot process and firmware on its Secure Enclave Processor, but, according to Martin, this is no more restrictive than modern PCs.
"In fact, mainstream x86 platforms are arguably more intrusive, as the proprietary UEFI firmware is allowed to steal the main CPU from the OS at any time via SMM interrupts, which is not the case on Apple Silicon Macs," he said.
More sales for Apple hardware is something they won't complain about. But they certainly won't help, and each new generation of Apple Silicon will probably be a new adventure for the people doing the porting. But even running on a year or two old Mac would be fine for Linus.
Apple sells around 18m laptops a year. Support for Linux would not increase sales by even a percentage point.
The attraction for Apple in allowing Linux (that is, allowing non-secure boot) would to be to avoid entanglement in accusations of monopolistic practices by US hardware and software manufacturers, practices the EU has been traditionally keen to prosecute and China may be increasingly keen on pursuing.
... to start off with grandiose statements such as This will include coding a driver for the "completely custom Apple GPU" , only to fail miserably, and then blame Apple for it?
Has Apple published the PRM for their GPU? If Apple doesn't publish the PRM for their custom GPU - which they will not - the only thing you'll get is something similar to the Nouveau NVIDIA driver on Linux. 11 years of development on Nouveau, and it's still an unusable pile of crap.
Same goes for power management.
Know thy limitations.
It is simply not possible to cleanroom reverse-engineer all the hardware features required to make this work well, within a reasonable timeframe, when the hardware manufacturer has spent untold resources on making their hardware as closed, obfuscated and impenetrable as humanly possible.
Not unless Apple decides to become a sharing benefactor of mankind and starts publishing the specs to everything. Which, so far, they have given no indication that they would want to.
This is Apple we're talking about here.
 Programmer's Reference Manual
Nouveau is actually brilliant. With it you can repurpose hundreds of machines, good enough to give to kids or office workers and simply run Linux and an office software. Much better than the landfill!
Out of any "innovation" I had to choose in the last 10 years, I would probably choose Nouveau. The other popular projects (Gnome 3, Wayland) so far have only *removed* useful features and damaged / fragmented the community.
Not everyone plays games or designs with CAD. For my personal use, I actually only need a framebuffer of a decent resolution. Granted, I won't buy NVIDIA hardware out of principle (same with Apple).
I've been doing that for years. Nouveau has literally never been one of my considerations. If I expect the user to be running Linux, office software, and a browser, I don't have to do very much to get extra graphics support. Sure, if they tried to stream 4K video to the old computer, it might have a problem, but usually it's old enough that it can't output that anyway. I don't need Nouveau to make LibreOffice usable; it will work fine even on ancient Intel integrated graphics. It's the more intensive stuff like image editing that needs a more poerful GPU, and as you've identified I usually don't expect the recycled computers to be used for that very often.
It’s usually not hard you just hook up the IO to a console and monitor the output and you manipulate the device, from there you can write a driver.
At least, that's what an moron on Macrumors told me (exact words quoted above) when I pointed out the difficulty of reverse engineering the custom Apple Silicon SOC to write Linux drivers. I didn't argue with him further, as he clearly had far more experience of being an idiot than I have.
"It’s usually not hard you just hook up the IO to a console and monitor the output and you manipulate the device, from there you can write a driver."
Which is why there has been so much success writing open drivers for systems like the RaspPi which also use ARM SoC's?
Getting a functional GPU driver is doable - getting a GPU driver that is able to exploit more than a tiny portion of a GPU's performance/functionality has proved to be very challenging indeed.
And as you approach a "workable" driver, the next generation of GPU or graphics standards arrive and the unfinished process for the previous driver slows and developers move onto the new hardware.
This is an attempt at injecting realism into the task rather than being overly critical - I am confident that they will have a significant level of success but there maybe better targets for that energy such as the Broadcom VideoCore because the market for Apple notebook users who want Linux and are prepared to cope with poor functionality in lots of areas isn't likely to be large when they can just use x86 equivalents.
It'll never work well enough to be a daily driver unless Apple help them.
So while I applaud the effort, they are doomed unless they can get some governmental entity to force Apple to help on pain of financial ruin.
And let's face it, the only way that happens is if the EU step in. The US aren't going to.
GalliumOS runs pretty nicely on a wide variety of (albeit older) Intel based Chromebooks. They don't (yet? ever?) support ARM devices.
In spite of their Linux-based underpinnings, I'm not aware of any other easy (or even difficult) ways to run a conventional Linux on any ARM based Chromebook, old or new. I'm hoping by the time of its demise in 2025, that there might be a viable offering for the Lenovo C330 (MTK8173C ARM based) and similar Chromebooks. If mine's still working by that point, I'd definitely switch it over.
If anyone is aware of any such Linux distribution for ARM based Chromebooks of any kind, I'm definitely interested.
It exists, but in each case it includes a bunch of system-specific cludges which might work. If you want to port, you can start with some of the OS code from the PineBook Pro, which is effectively the same kind of hardware as a Chromebook so more of their code is portable. Still, you'll have to get into the bootloader yourself, and the GPUs on ARM-based SOCs often lack driver support, which you'd have to fix. And then there's the drivers for other things, which should be standard but probably are not. So on second thought, you could always just buy a PineBook Pro if you want an ARM-powered light Linux laptop that you can be a little sure will work.
"I thought properly Linuxified Chromebooks were a thing. Maybe they're not powerful enough for Linus."
The most powerful one I've seen has six cores, 4 GB of RAM (maximum supported), and relatively slow access to storage. If you are hoping to do serious development on that of something large, compiling alone will take a very long time. If you want to use it because it's thin, light, and low-power, that works. If you want to develop small to medium projects on it, it will work but you'll notice a speed problem. If you want to develop a big project on it, you should have a good reason for not wanting to use anything else because you will have to sacrifice a lot of time and inefficiency. When compared to the Apple M1 chip, performance is not even comparable. It all comes down to the purposes to which the device will be put.
"His projects include porting Linux to the Sony PlayStation 4"
This is often quoted in the different press pieces about this.. but did you guys actually check what this port consisted of?
~40 mostly small commits to add a PS4 subarch to x86 and some small drivers that never got mainlined and has been abandoned since 2017.
And there's been zero discussion of how this will actually happen on either on LKML or the arm kernel list from what I can tell.
I can imagine something partially working (to console) in a few months and then you're looking at maybe years of release cycles if it does ever get mainlined due to all of the specifics needed to accommodate the Apple weirdness in the Linux arm64 support without making a mess.
I would like to be proven wrong but I think this might take long enough that AMD will have something that makes the M1 look like a toy and even Intel might have gotten their stuff together.
Why would anybody want to run Linux on an Apple computer? They are notoriously over-priced, and I don't suppose they would be much cheaper without the Apple operating system (even supposing that Apple allows them to be sold in such a way). One of the many attractions of Linux is that you can run it on nearly anything.
I can see the point in a Hackintosh - you get the nice Mac O/S on cheaper hardware. This project seems to be striving to put a free O/S on gratuitously expensive hardware.
Usually, people who want to run Linux on a Mac bought the Mac at least partially for Mac OS but also want to use Linux for other tasks. If you already like Apple software, running Linux on the hardware is still useful because you have the choice of both environments. At least it was when Linux could be installed without much hacking, maybe with a few less important hardware not having drivers, but mostly functional out of the box. It might not be so useful when it has to be custom untested code to work with Apple's new processors. If people buy Macs just to run Linux on them, they must really like the design of the laptops. I don't think that's often the intent.
Biting the hand that feeds IT © 1998–2021