igknighted said: I think you misunderstand OSX86. It is NOT running OSX on windows, it is running it on x86 hardware (as opposed to apple's PPC hardware, or specific x86 hardware). There is no real change to the kernel or anything like that. It is not running on a windows kernel as opposed to the mach (*NOT* BSD... but probably about as close to BSD as linux is) kernel. To change the kernel over would require a re-write of almost all the programs that interface it (most of them), and without the source code it would be impossible.
Besides, I am curious why linux would want a proprietary graphical interface when it already has incredibly competant free ones? What could possibly be good about it? Either it would make a small niche happy and not change the majority of users lives, or it would catch on and subvert the efforts of a lot of equally good (if not better) free software projects.
Pardon me, but I never said to run OSX on the NT kernel, I said PC arch in the meaning of generic x86, not apple specific hardware.
And yes there is kernel modification on the darwin core, there is some hacks to get informations from ACPI, vbe, and other bios specific firmware interfaces. SSE3 ISA interpreter for non-intel cpus (thus newly AMD don't need it) as I said before, the proprietary part intentionally depends upon EFI. So dummies EFI interfaces are generated for hardware check.
And what I said before, was creating an darwin ABI on linux would be very hard, and still might be unsuccessful, but if some one start the porting of darwin ABI now, it might be ready and stable in 2015 (Just take a look at the wine which started in 96).
Since the complexity from mach kernel, it would be easier to get the darwin and put linux as a modules.
Darwin is *NOT* *BSD but contains a BSD layer for networking, filesystem and POSIX compatibility.
It might be a good idea to create a linux layer for darwin...