I'm curious about three things: why a native-compileable pmax source
tree is so slow coming out, and what (if anything) I can do to help,
and if anyone is interested in the source patches I had.  I'd wanted
to not distribute them to anyone except the pmax port maintainer, but
if there's interest and no objection, I'll make the patches more
generally available.

Now that a native user-land exists, most of the ugly nasty hacks
necessary to get Ultrix binaries to run aren't necessary. The really
necesessary changes are relatively minor modifications to
pmax-specific files, mostly relating to a compiler that (unlike the
MIPSco toolset, which is what 4.4BSD more-or-less expected) prepends
underscores to external names.  That, and renaming all the .s files to
.S to make the Makefile rules coincide with GNU CC's suffix
conventions, and fixing those Makefile.inc that never knew about mips
or pmax targets to begin with.

These are all _really_ simple changes; the only issue is whether to
hide the _-or-not convention in a macro or just inline it.  I used a
macro, because that's trivial to change to an explict "_", and the
other way isn't.

If anyone is interested, what's a good format for these changes?
Context diffs from NetBSD-1.0, or something else? And one monster
context diff, or separate ones for the kernel and user code?

--Jonathan