Subject: Re: devfs, was Re: ptyfs fully working now...
To: Chapman Flack <>
From: Bill Studenmund <>
List: tech-kern
Date: 11/20/2004 17:25:42
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Sat, Nov 20, 2004 at 03:16:53PM -0500, Chapman Flack wrote:
> Bill Studenmund wrote:
> =20
> > representations. As we support more & more hot-swap devices, over time,=
> > may end up with multiple "devices" (say wedges/partitions on mobile dis=
> > with the same name. While obviously we can't have them active at the sa=
> > time, the fact that one wedge named "foo" was readable by user Bob does=
> > necessarily mean that every wedge named "foo" should be.
> This looks as if you're proposing an additional functional requirement, so
> let me make sure I understand it.  You would like a system where a single
> name in node-space can be configured to appear at different times with
> different ownership and/or permissions depending on what peripheral is
> plugged in?  I'm not at this point advocating or opposing such a feature,
> only making sure if that's what you meant.


I agree that this feature could be abused and lead to a real mess. However=
I think we need the base functionality of being able to cope with a device=
appearing that has/wants the same name as another device (where the other=
device either is or isn't currently present). If we use device name as a=20
primary key for devices, we will run into trouble.

Take care,


Content-Type: application/pgp-signature
Content-Disposition: inline

Version: GnuPG v1.2.3 (NetBSD)