Message-ID: <3BAB3018.CFE4290C@ieee.org>
Date: Fri, 21 Sep 2001 08:18:32 -0400
From: Bryan-TheBS-Smith <b.j.smith@ieee.org>
Organization: SmithConcepts/AbsoluteValueSystems
MIME-Version: 1.0
To: kris buggenhout <buggenkr@god.bel.alcatel.be>
CC: "linux-xfs@oss.sgi.com" <linux-xfs@oss.sgi.com>
Subject: Re: XFS to main kernel source
References: <Pine.LNX.4.33.0109211018430.21430-100000@helka> <3BAB119A.29888F3D@god.bel.alcatel.be>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-linux-xfs@oss.sgi.com
Precedence: bulk

kris buggenhout wrote:
> I dont think ext3 will be the contender as the design of ext3 is
> only a patched up ext2, with a lot of limitations to it...

Being that Ext3 is based on Ext2, a very trusted filesystem in many
circles, I'd reverse that statement "I don't think ext3 will be the
contender."  Even Hans Reiser has changed his opinion on Ext3,
because he finally realized that people are using it for those very
reasons.

> I would be more concerned if Reiserfs should get its act together or 

Although ReiserFS is a very novel and advanced filesystem, and is
probably the future of filesystem design, it's breaking of
traditional UNIX layout and structures is just too much of an issue
to those of us who run UNIX networks.  I need reliable NFS and I
need quota support.  Plus I don't trust its recovery mechanisms, as
its focus seems to be on features.

> IBM invests more time into Enterprise storage manager ( on top
> of jfs)

JFS seems to be lacking as much compatibility as ReiserFS. 
Something tells me that this is because it was a port from OS/2
rather than AIX's version???  I haven't used or tested JFS so I'm
making these statements blind.  Feel free to rebuke me on this, I
just looked at its feature list, what was missing and dismissed it
from any consideration.

> For the enterprose world I see Vxfs as a threat ... in unison with
> volume manager.If Veritas pulls that off in a reasonable time
> frame... a lot of company's will opt for Vxfs and volume manager,
> because it can be used on almost all of their platforms : wincrap2k,
> Solaris, Irix, Linux, HPUX, AIX,... etc...

I take it it is not OSS?  ;-PPP  It must be OSS or at least of
commodity cost to gain widespread acceptance, regardless of
superiority.

> having a unified fs architecture across platforms is a bonus ...
> knowledge of platform is not as important anymore regarding storage (
> which is a large part of systems management)....

True, oh so true.  Of course gaining access to Windows internals
isn't exactly something an OSS project can do.

-- TheBS

-- 
Bryan "TheBS" Smith   mailto:b.j.smith@ieee.org    chat:thebs413
Engineer  AbsoluteValue Systems, Inc.  http://www.linux-wlan.org
President    SmithConcepts, Inc.    http://www.SmithConcepts.com