From owner-freebsd-current Fri Jul 26 11:32:11 1996 Return-Path: owner-current Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id LAA14427 for current-outgoing; Fri, 26 Jul 1996 11:32:11 -0700 (PDT) Received: from brasil.moneng.mei.com (brasil.moneng.mei.com [151.186.109.160]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id LAA14415 for ; Fri, 26 Jul 1996 11:32:09 -0700 (PDT) Received: (from jgreco@localhost) by brasil.moneng.mei.com (8.7.Beta.1/8.7.Beta.1) id NAA17041; Fri, 26 Jul 1996 13:30:34 -0500 From: Joe Greco Message-Id: <199607261830.NAA17041@brasil.moneng.mei.com> Subject: Re: cvs commit: src/sys/dev/ccd ccd.c src/sys/dev/vn vn.c src/sys/sys conf.h src/sys/i386/isa fd.c mcd.c scd.c wcd.c wd.c wt.c s To: imp@village.org (Warner Losh) Date: Fri, 26 Jul 1996 13:30:33 -0500 (CDT) Cc: jkh@time.cdrom.com, freebsd-current@freebsd.org In-Reply-To: <199607261727.LAA28387@rover.village.org> from "Warner Losh" at Jul 26, 96 11:27:19 am X-Mailer: ELM [version 2.4 PL24] Content-Type: text Sender: owner-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > : It would also allow you to make links, like: > : > : ln -s /dev/cuaa0 /dev/mouse > : > : Which get recreated automagically when devfs is next mounted. Yes, > : you could accomplish the same effect with a handful of commands in > : /etc/rc.local, but the point is to make it transparent and so obey the > : principle of least astonishment. > > Hmmm, sounds like what is needed is a file system that maintains a > file, lets call it /etc/devperms to invent a name. When you mount > this file system, it reads that file. When you hack this file system, > it hacks that file for you. When you unmount, nothing would happen to > that file. All of this is dependent, of course, on the existance of > real devices on the system. I don't have anything useful to suggest other than to note that Sun took the easy way out of this problem by implementing a "devfs-on-ufs", i.e. they use UFS and have a procedure for "automatically" creating new instances of devices. Their method does preserve permissions, etc., but it does so at the awful price of a half-baked implementation that I really can't stand. ... JG