Subject: RE: suggested patch to console(4) manpage
To: Jonathan Stone <M.Drochner@fz-juelich.de>
From: Martin Husemann <martin@rumolt.teuto.de>
List: port-i386
Date: 01/20/2000 10:44:16
> What I would do (for now, until we can find out from Phoenix how to
> get at BIOS serial-console state) is turn on DIRECT_SERIAL.  Except
> that DIRECT_SERIAL is broken by design, and you're telling me I can't
> fix it.

No, that's not what he is telling you.
He told you: DIRECT_SERIAL does what I need. And it happens to be that way
because that was what I needed when I came up with the patches to implement
it; my old 486 with com0 connected to an original DEC VT420 that can't do
RTS/CTS but scrolling too slow to keep up with NetBSD booting at 9600 via
BIOS serial calls.

Matthias and I talked about implementing RTS/CTS at that time and decided
not to do it (a) because my terminal didn't understand it and (b) in his
setup the machine usualy boots without console attached. The possibility of
DOS by unpluging a console cable and causing some kernel printf scared us
further.

So the message was: please implement an option that switches DIRECT_CONSOLE
from XON/XOFF to RTS/CTS, but don't break the existing stuff.


Martin