lol my ADHD brain is a #RubeGoldberg [device], though. XD
Discussion
Loading...
CW: ARM kvetching
@rl_dane I really love ARM conceptually - it’s just stuck in a really bad loop because it only has one real mainstream adopter (Apple M-series like you mention) in the desktop world and said adopter locks everything down so hard…
As non-sensical of a pipe dream it is - I REALLY hope that Apple decides one day to just open source Rosetta (the system they made so that macOS’s x86_64 apps are still mostly backwards compatible), I think that has the possibility to give ARM platforms as a whole a true 180.
I also say this while I’m even more excited for anything to move in the world on RISC-V but that’s an even crazier dream that’ll never come true lol. Closest I’ll probably get (and something I’m considering acquiring) is the Framework RISC-V main board for some experimenting…
I don't see Apple open sourcing Rosetta, because they won't have much need for it in five years, maybe two. They were never much beholden to backwards-compatibility.
We need an ARM PC reference platform.
Heck, we need ARM embedded and mobile reference platforms!
Enough of this madness! XD
only has one real mainstream adopter in the desktop world
but that's half of the companies in the desktop world :)
@kabel42 @rl_dane True…I guess I mean more so in both the way of OS support and the fact that Intel and AMD control the other half of the hardware market and there’s no high compute ARM hardware I know of outside of M1.
Even if Windows or a popular Linux distro had fantastic support for it - there’s no commercially available ARM hardware that matches something like an Intel ultra 9 AFAIK
@roguefoam @rl_dane i was mostly thinking mainstream desktop is windows and mac but yes
Is winedows on are dead?
@roguefoam @rl_dane there was a windows on raspberry pi a few years ago, not sure that still exists.
And I thought there was a surface with ARM, but haven't heard a lot about that after release either.
1 more replies (not shown)
It's been on the back of my mind. ;)
The distance from back to front is a short trip. ;)
lol my ADHD brain is a #RubeGoldberg [device], though. XD
Same. Or #MouseTrap if you remember that game.
Apple sucks, but is that really the fault of ARM? Without Apple, you have open hardware interface standards, easy. I put the blame solidly on Apple for their hardware abuses.
Not that it matters, since nobody's doing anything serious with ARM other than Apple, but the Raspberry Pi is... fine?
Raspi is better than most ARM software by virtue of the fact that they value openness (to a limited extent), and that their hardware is very popular.
Apple, on the other hand, is extremely popular (as ARM goes, anyway), but they do not at all value openness, so you have herculean efforts like Asahi.
Other manufacturers like #Pine64 value openness (more so than Raspi), but they're not as popular.
The truth table ends up looking like this:
manu- | values | is | end
facturer | openness? | popular? | result
---------+-----------+----------+-------
RasPi | yes | yes | good
---------+-----------+----------+-------
Apple | NO | YES | meh
---------+-----------+----------+-------
Pine64 | YES | no | meh
---------+-----------+----------+-------
Yeah, just seemed odd to bring up the instruction set, when it's the SteveJobs virus responsible for making humans do this with it.
Well, typical dialectical minefield of me using a term to signify the entire platform and ecosystem, but that term theoretically only specifies the instruction set. 😅
As someone who has dealt with building firmware for ARM devices and dealing with all the poor SoC drivers, I understand your pain.
I'd like a universal boot environment. What chaps my trousers is that it already has one, but it isn't bothered with.
The ironic thing is that I'm about to deploy some embedded stuff, but may just go x86 because I know stuff like TPM controllers and other things work well enough, although I much prefer ARM.
CW: ARM kvetching
@rl_dane that's not an arm thing that's a not IBM/bios thing
If by "not an ARM thing," you mean that the ARM ISA wasn't designed to be intentionally opaque, I do agree, but given that that is how all of the hardware ends up, is the distinction useful?
Did you mean to say "not bios thing," or "that is a bios/ibm thing?"
@rl_dane if you replace arm with risc-v or MIPS you get the same problem. Not sure if PPC has a genetic solution?
I don't know much about MIPS, but I know RISC-V isn't the openness paradise that people assume it to be, just because the ISA is open. There's a lot more to hardware than just the ISA.
With PPC, there were at least some available specs for the hardware, and you had some kind of starting point. They had to for compatibility, so that Apple clone manufacturers could run MacOS 7, and PPC PCs could run Windows NT.
With ARM, every little piece of hardware is a unique, bespoke, hellsnowflake 😄
This essentially means though, that every single CPU architecture _except_ x86 will be worse than useless for all of time until anyone involved in them finally gets their head out of their ass and puts any level of effort what so ever into making their CPUs actually usable for anything but capitalist dystopias.
@OpenComputeDesign @rl_dane it's less x86 and more IBM-compatible or what you want to call it.
1+ more replies (not shown)
@OpenComputeDesign @rl_dane you could do all the things a BIOS or EFI does and use them in a computer with an ARM CPU. AFAIK Apple does that more or less. Just not with an accessible standard.
2+ more replies (not shown)
@rl_dane having hardware abstraction is a thing the bios does, you could have the same thing for arm and you can have x86 without
Yes, and x86_64 is slowly heading in that (awful) direction.
I don't see ARM ever moving in that (good) direction, but I'd be delighted to be proven wrong.
@rl_dane the latter actually exists, e.g. Newer playstation or xbox