Re: GGI, EGCS/PGCC, Kernel source

Michael Schmitz (
Sun, 1 Mar 1998 17:58:21 -0800

At 8:20 PM -0500 3/1/98, Theodore Y. Ts'o wrote:
> Date: Fri, 27 Feb 1998 15:00:18 -0800
> From: Michael Schmitz <>
> Let me try to put it this way: some machines (uClinux) have no keyboard,
> others want additional keyboards (multihead graphics independent of GGI).
> The serial lines seem to be a good choice. Specific serial lines are
> designated as keyboard input
>It's not at all clear this kind of routing of keycodes around should be
>done in the kernel --- indeed, I'll go farther, it almost certainly
>*shouldn't* be done in the kernel.
>There needs be basic functionality for booting a machine (the existing
>serial console and VT console functionality is a pretty good limit) and
>doing emergency system administration. However, supporting multiple
>keyboards for GUI applications (X windows or something else) should be
>done in user-mode, not in the kernel.

I'm sorry to have caused this confusion, I'm always lumping serial keyboard
and multiple keyboards together as I shouldn't...

Please note that your 'do this in user space' precludes two things: the use of
serial keyboards for machines that don't have keyboard but want to use the
video display for output (Palmpilot, that's how my interest in this topic
arose) and the use of VT consoles on multiple video displays with multiple
keyboards (not necessarily serial ones, think ADB or PS/2 here).
GUI applications running on multiple display heads sure can and should use
other methods to access their keyboards. But the world isn't all about X.


To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to