Subject: Re: /etc/inittab (was: /etc/default)
To: None <mrg@mame.mu.Oz.Au, woods@kuma.web.net>
From: Captech) <greywolf@tomcat.VAS.viewlogic.com (James Graham>
List: current-users
Date: 07/27/1995 09:35:08
#: From owner-current-users@NetBSD.ORG Wed Jul 26 20:29:42 1995
#: Date: Wed, 26 Jul 95 17:15:44 -0400 (EDT)
#: From: woods@kuma.web.net (Greg A. Woods)
#: To: matthew green <mrg@mame.mu.Oz.Au>
#: Cc: current-users@NetBSD.ORG
#: Subject: Re: /etc/inittab (was: /etc/default)
#: Reply-To: woods@kuma.web.net (Greg A. Woods)
#: Organization: Planix, Inc.; Toronto, Ontario; Canada
#: 
#: [ On Wed, July 26, 1995 at 16:45:37 (+1000), matthew green wrote: ]
#: > Subject: Re: /etc/default 
#: > 
#: > as much as most people seem to hate it, AIX lets you have both
#: > sysv /etc/init.d and bsd /etc/rc* init scripts.  can we steal
#: > from here and make everyone happy?
#: 
#: The real key to the puzzle is an init(8) with run-levels and something
#: like /etc/inittab.  From there anything is possible.  That's what
#: AIX-and-pains (and old NCR SysV, etc) does...
#: 

In other words, SVRx init.

I think the general consensus has been trying to find a way around this.
How many "run levels" do you _need_?  I can only think of four:

	halted
	single-user
	multi-user standalone
	multi-user networked (or, as I put it earlier, "multi-user with
		a side of bacon")

The last two may seem redundant but if you have a machine you're migrating
on and off the network, it might be nice to have a mechanism in place to
facilitate this.  Since we're all resourceful hackers to some extent,
everyone will probably come up with their own way of doing things.
rc.net was a good step in this direction;  we have only to make use of
a switch which avoids the rc.net file and make sure that we have standalone
functionality contained entirely outside rc.net.

				--*greywolf;
	
#: -- 
#: 							Greg A. Woods
#: 
#: +1 416 443-1734			VE3TCP			robohack!woods
#: Planix, Inc. <woods@planix.com>; Secrets Of The Weird <woods@weird.com>
#: