From owner-freebsd-hackers  Fri Dec 12 17:41:13 1997
Return-Path: <owner-freebsd-hackers>
Received: (from root@localhost)
          by hub.freebsd.org (8.8.7/8.8.7) id RAA14386
          for hackers-outgoing; Fri, 12 Dec 1997 17:41:13 -0800 (PST)
          (envelope-from owner-freebsd-hackers)
Received: from kai.communique.net (Kai.communique.net [204.27.67.90])
          by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id RAA14345
          for <freebsd-hackers@freebsd.org>; Fri, 12 Dec 1997 17:41:03 -0800 (PST)
          (envelope-from nectar@kai.communique.net)
Received: (from smap@localhost)
	by kai.communique.net (8.8.8/8.8.7) id TAA00563;
	Fri, 12 Dec 1997 19:41:08 -0600 (CST)
Message-Id: <199712130141.TAA00563@kai.communique.net>
X-Authentication-Warning: kai.communique.net: smap set sender to <nectar@kai.communique.net> using -f
Received: from localhost.communique.net(127.0.0.1) by kai.communique.net via smap (V2.0)
	id xma000554; Fri, 12 Dec 97 19:40:48 -0600
From: Jacques Vidrine <n@nectar.com>
To: Terry Lambert <tlambert@primenet.com>
cc: n@nectar.com (Jacques Vidrine), dnelson@emsphone.com, j_mini@efn.org,
        AdamT@smginc.com, jkh@time.cdrom.com, freebsd-hackers@freebsd.org
Subject: Re: Kernel Config datafile... 
In-reply-to: <199712130041.RAA19766@usr04.primenet.com> 
References: <199712130041.RAA19766@usr04.primenet.com>
Date: Fri, 12 Dec 1997 19:40:48 -0600
Sender: owner-freebsd-hackers@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk


Tee hee.  

Your later suggestion (by later I mean the one I read after I wrote
the message below), if I read it right, sounds like the way to go long
term.  I understood it to mean something to the effect that at kernel
build time, device drivers should probe the hardware to decide for
themselves whether they should be included in the linked kernel.

Or perhaps you meant that the kernel should be finalized at system
load time via dynamic loading, rather than during kernel build time.

I knew I should have read it twice before deleting, maybe you can
privately resend?

I guess the question it also brings up is, Do we want to spend time on
an intermediate solution that _does_ involve a configuration file?  Or
do we want to skip reinventing the "Windows registry" and go for
reinventing "Plug-n-Play" :-)

Jacques Vidrine <n@nectar.com>

On 13 December 1997 at 0:41, Terry Lambert <tlambert@primenet.com> wrote:
> > sio:
> >   parent = isa?
> >   port = 0x3f8 "COM1", 0x2f8 "COM2", 0x3e8 "COM3", 0x2e8 "COM4", \
> >          integer "OTHER"
> >   tty = true
> >   flags = 0x00001 "Shared IRQs", 0x00002 "Disable FIFO", ...
> >   irq = 3 "COM2 or COM4", 4 "COM1 or COM3", any "OTHER"
> >   vector = siointr
> 
> Instead of reinventing Windows 3.1 .INI files, why down't we just
> jump straight to reinventing the the Windows 95 registry?
> 
> 
> 					Terry Lambert
> 					terry@lambert.org
> ---
> Any opinions in this posting are my own and not those of my present
> or previous employers.