Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 25 Oct 2002 11:05:04 -0700 (PDT)
From:      Julian Elischer <julian@elischer.org>
To:        Mark Valentine <mark@thuvia.demon.co.uk>
Cc:        Poul-Henning Kamp <phk@critter.freebsd.dk>, "M. Warner Losh" <imp@bsdimp.com>, freebsd-arch@freebsd.org
Subject:   Re: cvs commit: src/lib/libdisk Makefile chunk.c write_alpha_disk.c write_i386_disk.c write_pc98_disk.c
Message-ID:  <Pine.BSF.4.21.0210251041050.6512-100000@InterJet.elischer.org>
In-Reply-To: <200210251735.g9PHZ1sT075954@dotar.thuvia.org>

next in thread | previous in thread | raw e-mail | index | archive | help


On Fri, 25 Oct 2002, Mark Valentine wrote:
> 
> I'm sure we'll appreciate the progress you've made once we've seen what
> it gives us.  We're simply asking why we have to sacrifice a traditional
> interface (and in my experience one which is more reliable) to buy into it.
> 
> What is the cost of having both?

You can't have both because the naming scheme gives a particular name to 
a partition because of how you get to it..
ad0a means:
I openned the disk, found a disklabel and this is the first partition
defined in that disklabel.

We've been (except for Bruce, (often right, but not one to part with old
ways)) moving towards doing this for the  last 10 years..
Devfs and it's associated disk-recarving code have been postulated and
prototyped singe 1993. We have made the default disk names installed
the "new, reality based" ones since about 1996. If someone doesn't know
what ad0s1a means yet then they haven't been paying attention.

As new devices come and go in the new 'pluggable, wireless world' the
old scheme of assigning for all eternity a particular set of limitted
availability bits to a particular drive and part of that drive, has to
give way to something else.  Devfs is one possibility for this, but
devfs cannot easily cope with the multiple ways that drive partitionning
can be layered, without breaking away from the traditional fixed
geometry<->minor-number/naming scheme. In the old scheme there is just
NO WAY to describe some partitioning schemes, especially in the plug-in
world where the drive may have been partitioned by some medical imaging
system or a MAC or a SUN. Our hand is being forced. 

<dreaming>
 In the future I feel that the default names may eventually be
over-ridden by semantically meaningfull names. e.g. You may partition a
drive and give each partition a name which might be used instead of the
default inherrited name.. e.g. ad0s1s2b (not a typo) may decide that it
would rather be known "/dev/swap2" because the table entry has a field
"swap%d" in it, or disk2 may decide that it wants to be known as
"/dev/tunes". (A removable disk full of music)

These things are possible with devfs.

/dev/tunes-s1a would be a partition on that drive.. it's removable
so when you replace it with the drive that calls itself "/dev/backups"
the /dev/entries correct themselves. But that is all just potential
at this stage. 
</dreaming>

To address some of your concerns
We COULD have a /dev/ad0sBa that always reflects the first BSD slice
"a" partition. that would give the characteristics you have asked for..
and still abides by the naming convention.. (devfs could make a symlink
or something..) but ad0a is already taken..




To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-arch" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.21.0210251041050.6512-100000>