From: 
   Date: Fri, 06 Jan 1995 02:12:17 -0700
   Sender: hpeyerl@sidney.novatel.ca
   
   
   Jonathan;
   
   I'm curious.  What does the version info at the top of your pmax/pmax/cons.c
   say?

Okay.  I'm not naming names to assign blame but to try and reflect
what happened here.

Christos made a change to support device  cloning and changed
all th device  open code.  He missed the pmax console.  When
I found that by noticing my -current kernel had a type error,
I griped at him and he applie my fix -- which was to pass
a clone argument to the pmax console open routine, and pass
that on to any subsequent lower-level open routine.   When someone _else_
backed that change out, they _also_ forgot to back it out of the pmax port.

It seems like no-one outside the pmax port knows that the pmax
has its own console code.  That's why I raised the idea of doing
away with as much of the pmax console code as possible, and
using src/sys/dev/cons.c instead.

I hypothesise (read: consed out of the ether) that pmax/cons.c only
continues to exist because it's in src/sys/arch/pmax/pmax, not in
src/sys/arch/pmax/dev.