4 ms·

> main devs have moved on How confident are you in this statement? I have no particular knowledge of Asahi. But I do know this narrative emerged about Rust-for

by bjackman 10mo ago

> main devs have moved on How confident are you in this statement? I have no particular knowledge of Asahi. But I do know this narrative emerged about Rust-for-Linux after a couple of high-profile individuals quit. In that case it was plainly bogus but this was only obvious if you were somewhat adjacent to the relevant community. So now I'm curious if it could be the same thing. (Hopefully by now it's clear to everyone that R4L is a healthy project, since the official announcement that Rust is no longer "experimental" in the kernel tree). I know Asahi is a much smaller project than R4L so it's naturally at higher risk of losing momentum. I would really love Asahi to succeed. I recently bought a Framework and, while I am pretty happy with it in isolation... when I use my partner's M4 Macbook Air I just think... damn. The quality of this thing is head and shoulders above the rest of the field. And it doesn't even cost more than the competition. If you could run Linux on it, it would be completely insane to use anything else.

neobrain 10mo ago
It's similarly bogus here. Early Asahi development tried to upstream as much as possible but ultimately still maintained a gigantic pile of downsteam patches, which wasn't a sustainable model. Most of current development is focused on reducing that pile to zero to get things into a tractable state again. So things continue to be active, but the progress has become much less visible.
- adastra22 10mo ago
  M2 to M3 was a complete architectural change that will require a lot of reverse engineering. As far as I know no one is working on this. The M1/M2 work was a labor of love of largely one dev that has since moved on. The project is still active and working to upstream the work of these devs. But as far as I know, no NEW reverse engineering is being done. Ergo, it’s a dead end. Would be happy to be proven wrong.
TimTheTinker 10mo ago
Someone should create a minimal, nearly-headless macOS distribution (similar to the old hackintosh distros) that bootstraps just enough to manage the machine's hardware, with no UI, and fires up the Apple virtualization framework and a Linux VM, which would own the whole display.

neobrain 10mo ago

It's similarly bogus here. Early Asahi development tried to upstream as much as possible but ultimately still maintained a gigantic pile of downsteam patches, which wasn't a sustainable model. Most of current development is focused on reducing that pile to zero to get things into a tractable state again. So things continue to be active, but the progress has become much less visible.

adastra22 10mo ago
M2 to M3 was a complete architectural change that will require a lot of reverse engineering. As far as I know no one is working on this. The M1/M2 work was a labor of love of largely one dev that has since moved on. The project is still active and working to upstream the work of these devs. But as far as I know, no NEW reverse engineering is being done. Ergo, it’s a dead end. Would be happy to be proven wrong.

TimTheTinker 10mo ago

Someone should create a minimal, nearly-headless macOS distribution (similar to the old hackintosh distros) that bootstraps just enough to manage the machine's hardware, with no UI, and fires up the Apple virtualization framework and a Linux VM, which would own the whole display.