From owner-cvs-all Fri Feb 13 15:48:06 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id PAA06121 for cvs-all-outgoing; Fri, 13 Feb 1998 15:48:06 -0800 (PST) (envelope-from owner-cvs-all@FreeBSD.ORG) Received: from dingo.cdrom.com (dingo.cdrom.com [204.216.28.145]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id PAA05821; Fri, 13 Feb 1998 15:45:37 -0800 (PST) (envelope-from mike@dingo.cdrom.com) Received: from dingo.cdrom.com (localhost [127.0.0.1]) by dingo.cdrom.com (8.8.8/8.8.5) with ESMTP id PAA05222; Fri, 13 Feb 1998 15:43:17 -0800 (PST) Message-Id: <199802132343.PAA05222@dingo.cdrom.com> X-Mailer: exmh version 2.0zeta 7/24/97 To: Poul-Henning Kamp cc: Mike Smith , Julian Elischer , Paul Traina , core@FreeBSD.ORG, junichi@jp.freebsd.org, committers@FreeBSD.ORG Subject: Re: wfd block major number reassignment from 24 to 1 In-reply-to: Your message of "Sat, 14 Feb 1998 00:21:37 +0100." <3663.887412097@critter.freebsd.dk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Fri, 13 Feb 1998 15:43:16 -0800 From: Mike Smith Sender: owner-cvs-all@FreeBSD.ORG Precedence: bulk > In message <199802132319.PAA05082@dingo.cdrom.com>, Mike Smith writes: > >> I belive that persistence in DEVFS is a BAD thing, but I'm appearantly > >> pretty alone on -core with this view... > > > >Have you ever wanted to change permissions on an entry in /dev? > > yes. Ok. > In which case I always add it to /etc/rc.local so I'm sure it will > be there on the next reboot. Why? Do you do this for all the other files throughout your system that you change the permissions on? Do you *expect* permissions on arbitrary files to change when you reboot? > >If not, then your stance is understandable. But as soon as you accept > >that there may be more than one "right" set of permissions for > >something, you accept that persistence is required. > > Yes I have, and no I do not accept that. But you do. Your entry in /etc/rc.local is an implementation of persistence. Unfortunately, it's not adequte for at least the following reasons: - If you mount devfs somewhere else (think "chroot()"), the new permissions are not necessarily going to come across (cf. conversations with Julian on this). - There is no logical connection between /etc/rc.local and /dev - If a device node first appears after /etc/rc.local is run, it is not possible to set its permissions. This is an issue for PCI, PnP, PCCARDs and CARDBUS at the very least. > If I decide that disks should be "642 foo.mumble" on my machine, > I should be able to express that such that when I add more disks > it will DTRT. > > Permissions in /dev is a policy issue, and should be handled as > such: ie, from a root-controlled config file. No argument with anything there. > I belive persistence in devfs (as in /dev) is a very bad thing. You certainly haven't expressed that. In fact, everything else you have said in this message directly contradicts that. I can only conclude that you're reading a different meaning into "persistence" than I am, and by inference a lot of other people. -- \\ Sometimes you're ahead, \\ Mike Smith \\ sometimes you're behind. \\ mike@smith.net.au \\ The race is long, and in the \\ msmith@freebsd.org \\ end it's only with yourself. \\ msmith@cdrom.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message