Re: no need for a devfs

Kai Henningsen (kaih@khms.westfalen.de)
11 Jan 1998 22:56:00 +0200


mingo@chiara.csoma.elte.hu (MOLNAR Ingo) wrote on 11.01.98 in <Pine.LNX.3.96.980111202932.15060B-100000@chiara.csoma.elte.hu>:

> On 10 Jan 1998, Kai Henningsen wrote:
>
> > pbrutsch@creighton.edu (Phil Brutsche) wrote on 07.01.98 in
> > <Pine.HPP.3.95.980107213351.13482C-100000@bluejay.creighton.edu>:
> >
> > > No, please don't do this. It's too confusing. What's wrong with the
> > > current SCSI setup (/dev/sda, /dev/sda1, /dev/sda2, etc) in any case?
> >
> > It's fragile. But then, so is the proposed replacement.
>
> ? whats the problem with devfs-v3? It removes the dependence on _any_
> numbering scheme. You pick a logical naming (in string-space) _once_, and
> it will be supported forever. The mapping between devices and internal
> numbering (major/minor) is transparent.

That's true for all these schemes.

The important question is, what does the name _mean_? With the current
scheme, "/dev/sdb7" means "the 3rd logical partition on the 2nd SCSI
disk". "/dev/hdb7" is a little better - "the 3rd logical partition on the
primary IDE slave disk".

What I proposed was something meaning effectively "the filesystem I mkfs'd
on 1997-12-17 13:55:21.765".

See the difference?

MfG Kai