Subject: What about old talk protocol?
To: None <ragge@ludd.luth.se>
From: Gordon W. Ross <gwr@mc.com>
List: current-users
Date: 10/19/1995 16:52:00
> From: Anders Magnusson <ragge@ludd.luth.se>
> Date: Thu, 19 Oct 1995 17:39:02 +0100 (MET)

> > Unfortunately, Sun still uses the old talk protocol, which I have been told
> > is broken in a couple of ways. I don't think the old 'talk' was encumbered,
> > so you could still get it from an old BSD distribution, but I am
> > not sure. Or have a look at ytalk, which should do both protocols.
> > 
> You also have another problem: Old talk protocol is byte-order dependent.
> This means you can't ever talk to a Sun machine from a little-endian
> machine.

Now that "ntalk" is available, would it help to pretend that the
"old talk" protocol was specified to use network-order numbers?
At least then Sun's and m68k boxes would work with NetBSD (well,
better than they do now, anyway) and little-endian machines could
just use "ntalk" instead.  Better still, maybe there is some
heuristic that could be used to guess your peer's byte order.

Just some ideas...