Tom Uban wrote:
Perhaps there is no definitive CP/M machine because
CP/M was the
equivalent of an early unix/linux where it was possible to run the
code (or an easy modification thereof) on so many platforms that
it can not be locked to any one in particular.
The Intel MDS-800 was the canonical CP/M machine, in that it was the
only machine directly supported by Digital Research, at least for 1.x
and 2.x.
That's certainly not saying that it was best, or definitive (whatever
that means).