From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932323AbWFJDjK (ORCPT ); Fri, 9 Jun 2006 23:39:10 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932328AbWFJDjK (ORCPT ); Fri, 9 Jun 2006 23:39:10 -0400 Received: from fmr18.intel.com ([134.134.136.17]:47840 "EHLO orsfmr003.jf.intel.com") by vger.kernel.org with ESMTP id S932283AbWFJDjI (ORCPT ); Fri, 9 Jun 2006 23:39:08 -0400 Date: Fri, 9 Jun 2006 20:37:36 -0700 From: Valerie Henson To: Andrew Morton , Sonny Rao , jeff@garzik.org, hch@infradead.org, cmm@us.ibm.com, linux-kernel@vger.kernel.org, ext2-devel@lists.sourceforge.net, linux-fsdevel@vger.kernel.org Subject: Re: [Ext2-devel] [RFC 0/13] extents and 48bit ext3 Message-ID: <20060610033735.GH10524@goober> References: <1149816055.4066.60.camel@dyn9047017069.beaverton.ibm.com> <20060609091327.GA3679@infradead.org> <20060609030759.48cd17a0.akpm@osdl.org> <44899653.1020007@garzik.org> <20060609095620.22326f9d.akpm@osdl.org> <4489AAD9.80806@garzik.org> <20060609103543.52c00c62.akpm@osdl.org> <20060609214200.GA18213@kevlar.burdell.org> <20060609151553.30097b44.akpm@osdl.org> <20060609231151.GL5964@schatzie.adilger.int> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20060609231151.GL5964@schatzie.adilger.int> User-Agent: Mutt/1.5.9i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Fri, Jun 09, 2006 at 05:11:52PM -0600, Andreas Dilger wrote: > On Jun 09, 2006 15:15 -0700, Andrew Morton wrote: > > > > We're continuing to nurse along a few basically-15-year-old filesystems > > while we do have the brains, manpower and processes to implement a new, > > really great one. > > > > It's just this feeling I have ;) > > I think many people share this feeling (me included), hence the linux > filesystem meeting next week... The problem is that even getting a > half-decent disk filesystem is many years of work, and large disks are > here before then. The ZFS code took 10 years to get to its current state, > I understand, so I don't anticipate we will get there overnight. I helped bring up the first instance of ZFS running as a kernel module on Halloween, 2002 (one fun week staying up all night hacking with Jeff Bonwick). The earliest code was written in either 2001 or just possibly 2000 - so 5-6 years in elapsed time. On the other hand, in terms of total programmer staff-years put into ZFS, it's on the order of 25 years. I'm not sure either what the best route to the next big Linux file system is - start from scratch or reuse a lot of code. One of the things I want to talk about at the workshop is creative reuse of existing code, a la the continuation inode idea. -VAL