Comment by saagarjha
15 hours ago
I assume the idea is that finding tools and assembling other projects together into a build environment is comparatively easy but papering over entire components being missing is much harder
15 hours ago
I assume the idea is that finding tools and assembling other projects together into a build environment is comparatively easy but papering over entire components being missing is much harder
No. As I said, that's really a secondary niggle with the paper itself misrepresenting its reference, which as you can see already provides patches to paper over such stuff, as through the problem with XNU were redaction, which it isn't per that reference.
The primary reason not to use XNU is what the paper goes into in detail; which is the architecture of XNU simply getting in the way, just as the architecture of NetBSD would for the same reasons. If XNU being incomplete were the primary problem, NetBSD, a complete operating system that supports loadable kernel modules and provides a coherent development toolchain out of the box, would be the answer. But it is not.
Just because someone did it doesn't make it easy? I found the reference to support the claim fairly well