From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1030552AbWFIV6r (ORCPT ); Fri, 9 Jun 2006 17:58:47 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1030553AbWFIV6r (ORCPT ); Fri, 9 Jun 2006 17:58:47 -0400 Received: from terminus.zytor.com ([192.83.249.54]:51918 "EHLO terminus.zytor.com") by vger.kernel.org with ESMTP id S1030552AbWFIV6q (ORCPT ); Fri, 9 Jun 2006 17:58:46 -0400 Message-ID: <4489EF01.30001@zytor.com> Date: Fri, 09 Jun 2006 14:58:25 -0700 From: "H. Peter Anvin" User-Agent: Thunderbird 1.5.0.2 (X11/20060501) MIME-Version: 1.0 To: Jeff Garzik CC: linux-kernel@vger.kernel.org Subject: Re: [Ext2-devel] [RFC 0/13] extents and 48bit ext3 References: <4489D36C.3010000@garzik.org> <20060609203523.GE10524@thunk.org> <4489EAFE.6090303@garzik.org> <4489EEA7.8010704@garzik.org> In-Reply-To: <4489EEA7.8010704@garzik.org> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Jeff Garzik wrote: > > Right, and that proves my point. When you start making major changes > like 32->64 bit block numbers, you should communicate to the user (with > a big blinky "ext4" sign) that his filesystem metadata will change a > lot, not a little. Not to mention that such code will add yet more "if > (new) .. else .." code. > Forking the code and registering another filesystem name are two separate things, though. You can easily have a single filesystem module register two different filesystem names. -hpa