Stathol’s post is a perfect detailed description of why the mouse moves. Excellent write-up!
Is that true or is it derived from Mach 2.5 which did not have the OS running in user space on top of a microkernel, but rather a more traditional architecture?
Well, I don’t know if there’s really a clear delineation between which versions Mac OS did or didn’t draw from. It might be more accurate to say that it was “inspired” by Mach more than “derived” from it in any technical sense. I don’t know to what extent any code from Mach (of any version) was actually used, modified or not, in Apple’s OS designs. In contrast, I’m fairly sure that parts of the OpenBSD/NetBSD code base were actually used (some directly, some modified) in the predecessors of Mac OS.
That said, the gist of your objection is basically correct. I probably should have just left it at “the Mach kernel” without reference to version numbers. Mac OS X may have drawn on Mach 3 in certain areas, but overall, it is more similar to Mach 2.5. In retrospect, it looks like I’m claiming that Mac OS is a microkernel design. It isn’t, and I didn’t mean to imply that it was. I only meant to say that it was inspired by microkernel principles. Mac OS, like Mach 2.5, is a hybrid kernel that bridges the gap between microkernel and macrokernel. The OS spectrum is something like:
(macrokernel) Windows – Linux – Mac OS, Mach 2.5, BSD – Mach 3 (microkernel)