|
|
|
|
|
by jokethrowaway
1467 days ago
|
|
Because it's a hell of a task, regardless if the outcome will be open or closed. The closest thing I can think of is Apple taking BSD and making Darwin. Which company would be happy to risk so much, in such a busy market as mobile OS? |
|
Disclaimer: i'm no hardware engineer
I believe this is utterly false. There is a strong distinction between selling open hardware (the real stuff, not just open board design) and selling closed hardware.
If all you care about is board schematics, it's not hard to buy various chips and assemble them into functioning hardware. It's then often (though not always) possible to use the various binary firmware and drivers provided by the vendors for a certain platform (say Android device tree, or Windows driver) and make the hardware usable under a specific system.
Now, if you're trying to do real open hardware and have 100% schematics available, good luck: it's technically possible to acquire an open hardware micro-controller/CPU (though you have to look for it), but then you need open hardware GPU and network card and i don't think that even exists (at least not fulfilling our 2022 feature expectations).
These problems are detailed in various sources. My two favorites are an article about Lenovo hardware support in HaikuOS and a recent talk about operating systems design for interfacing with hardware:
https://www.haiku-os.org/blog/mmu_man/2021-10-04_ok_lenovo_w...
https://www.usenix.org/conference/osdi21/presentation/fri-ke...