Date:        Sun, 22 Mar 98 15:27:06 -0500 (EST)
    From:        woods@kuma.web.net (Greg A. Woods)
    Message-ID:  <m0yGrKk-0009wFC@kuma.web.net>

  | Indeed, however in the past most such problems happen with commercial
  | systems for which few users have driver source code, let alone full OS
  | source code.

Note I didn't say who should fix the bugs... 

  | Even worse is the situation where very few OS vendors have people smart
  | enough to work out the logic necessary

I actually doubt that, most of them have several very smart people.  The
problem is that working on tty drivers is not the sexy end of the todo list,
and the good people can usually find something more attractive to interest
them...

  | Actually it's not quite that simple.  99.9% of competent system
  | administrators won't change the naming schemes in /dev even if they
  | think they can.

Sure, no point playing with what works, the point is just that if you
really hate the naming that has been foisted upon you, you can change it.

  | Such magic can be hidden in well designed and documented library
  | routines such that any old program can be built to use dialout ports.

No, this is simply wrong.   For devices where it makes no sense at all
to use them without extra magic support of some kind (like network interfaces
for example, where addressing, etc, all needs to be grafted in somehow)
that's some argument for this, but for any device where simply opening it
and sending data in a stream, which certainly includes all tty type devices
(and parallel ports, disks, tapes, ...) it ought be possible to use any of
those interchangably in the same program without the program needing to know
anything about where its output is going (ie: I ought be able to
	:w /dev/serial-line
from in vi - and vi must not have to learn how to open this magic device).

This is one of the unix fundamentals - taken even further in plan9 of
course - that all devices look just like files, it is one of the attributes
that helped unix succeed.

kre