From edsews!uunet!fsg!skipnyc!skipsun!skip Fri Feb 2 14:00:31 1990
Received: by edstip.eds.com (3.2)
id AA20067; Fri, 2 Feb 90 14:00:29 EST
Received: by edsews.EDS.COM (smail2.5)
id AA26754; 2 Feb 90 04:47:33 EST (Fri)
Received: from fsg.UUCP by uunet.uu.net (5.61/1.14) with UUCP
id AA20948; Thu, 1 Feb 90 05:18:47 -0500
Received: from skipnyc.UUCP by mapes.fsg.com (4.0/SMI-4.0)
id AA20556; Thu, 1 Feb 90 03:54:44 EST
Received: by skipnyc.UUCP (smail2.5)
id AA00370; 1 Feb 90 03:51:14 EST (Thu)
Received: by skipsun (4.0/SMI-4.0)
id AA25325; Thu, 1 Feb 90 01:56:41 EST
From: uunet!fsg!skipsun!skip (Skip Gilbrech)
Message-Id: <[email protected]>
Subject: Re: Xenix Faster Keyboard repeat wanted.
To: edsews!edstip!ohrnb (Erik Ohrnberger)
Date: Thu, 1 Feb 90 1:56:38 EDT
In-Reply-To: <[email protected]>; from "Erik Ohrnberger" at Jan 18, 90 5:07 pm
X-Mailer: ELM [version 2.2 PL9]
>Is there anyway to increase the keyboard repeat rate
>on a 386 running Xenix? Under MS/PC-DOS I seem to recall
>a program in C that "outb" some instructions to the
>keyboard controller that told it to increase its
>repeat rate. Can this be used while running Xenix?
>Does anyone know of the command sequence to the
>controller? Does anyone have a smattering of C
>code that does this?
I've been using the following code since I got 286 Xenix long, long ago
(hasn't destroyed anything yet..). It's not terribly flexible, robust,
or pretty (my god, I just noticed I used wait loops!!), but you only need
to invoke it once at system startup time, and it makes the keyboard much
more pleasant to use..
By the way, you need '/dev/port' to use this, and as I recall, this file
doesn't exist in the system as distributed, even though the kernel supports
it. At least in 386 Xenix 2.2.6 it's major device 4, minor 3, so you would do:
% /etc/mknod /dev/port c 4 3
Unless your machine is totally isolated/personal, I'd also set the
permissions very carefully, since you can obviously really screw up
the machine by writing to arbitrary ports...
------------------------------ cut here ------------------------------