> It won't get better than this,

Microsoft don't let you access perf_event, which immediately makes it useless for me for example.

It's not really fragmentation as much as there not being being a billion dollar company making the stack actually work in the first place - if Apple decided to rebase onto Linux for some reason there's nothing stopping them getting everything shipshape just that it takes a lot of time and money. The vast majority of Linux boxes distro's are probably pareto distributed so 20-80 to make up a random number. The stack isn't that sparse.

If Linux decided to switch to a different license, like say a BSD-style that would allow Apple to do the things they want to do commercially, then Apple might consider that option.

However, the probability that Linux will switch to a license that is sufficiently Apple-friendly is precisely as likely as Apple switching the core of their OS over to Linux.

In other words, approximately 0.00000000000000000000000000000000000%.

I said "for some reason" to indicate that Apple would never do this, not least because of NIH.

Apple will never bother again with GPL software, while Google would rather migrate everyone to Fuchsia if they could, IBM is focused on making Linux a good citizen for cloud and mainframe computing, Oracle just wants to make it run their database stack, so that leaves out Microsoft, in what concerns billion dollar companies that would care to do that.

That has nothing to do with what I said.

Sure it has, Apple has no reason to adopt Linux on their consumer devices, in fact they are slowly moving away from POSIX, e.g. new networking stack.

And then I gave the list of the companies with very deep pockets that have any reason to ever bother having a Desktop Linux of some form.