Date: Fri, 26 Jul 1996 17:32:27 -0700 From: David Greenman <dg@root.com> To: Poul-Henning Kamp <phk@critter.tfs.com> Cc: Joe Greco <jgreco@brasil.moneng.mei.com>, wollman@lcs.mit.edu (Garrett Wollman), current@FreeBSD.ORG, imp@village.org, jkh@time.cdrom.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 Message-ID: <199607270032.RAA23533@root.com> In-Reply-To: Your message of "Fri, 26 Jul 1996 21:37:22 %2B0200." <3119.838409842@critter.tfs.com>
next in thread | previous in thread | raw e-mail | index | archive | help
> >As much as I pride myself with never agreeing with Garrett, I'm having a >hard time finding a reason not to in this particular case. :-) > >There are a couple of fine points still not resolved. >There is >still< no way to generally specify a policy: > "Make all disks foo.bar mode 0640" >and such. > >The major/minor of the underlying nodes are of course of no significance, >but wouldn't it be nice to get rid of all that blasted code entirely ? > >Why would we even care about the type of node apart from "Directory" or >"Not Directory" ? Ie, we could mount devfs on a FS that doesn't support >c/b devs at all (msdos, hfs ...) I tend to agree with Poul's opinion...I'd really like to get rid of c/b devices entirely (in as far as they are types within the filesystem). Only devfs should know about them. Perhaps the underlying file in Garrett's suggestion can be of any type - even a regular file of 0 bytes. The new instances that he talks about being created automatically would be 0 length regular files. If we go for a model like this, I would want it to be controlled with a (devfs) mount option so that persistence can be turned on and off. -DG David Greenman Core-team/Principal Architect, The FreeBSD Project
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199607270032.RAA23533>