To be clear, the work the asahi folks are doing is incredible. I’m ashamed to say sometimes their documentation is better than the internal stuff.

I’ve heard it’s mostly because there wasn’t an m3 Mac mini which is a much easier target for CI since it isn’t a portable. Also, there have been a ton of hardware changes internally between M2 and M3. M4 is a similar leap. More coprocessors, more security features, etc.

For example, PPL was replaced by SPTM and all the exclave magic.

https://randomaugustine.medium.com/on-apple-exclaves-d683a2c...

As always, opinions are my own

This is what ruffles my jimmies about this whole thing:

> I’m ashamed to say sometimes their documentation is better than the internal stuff.

The reverse engineering is a monumental effort, this Sisyphean task of trying to keep up with never-ending changes to the hardware. Meanwhile, the documentation is just sitting there in Cupertino. An enormous waste of time and effort from some of the most skilled people in the industry. Well, maybe not so much anymore since a bunch of them left.

I really hope this ends up biting Apple in the ass instead of protecting whatever market share they are guarding here.

I strongly support a projects stance that you shouldn't ask when it will be done. But the time between the M1 launch and a good experience was less than the time since M3 I would love to know what is involved.