Received: from spf3.us4.outblaze.com (spf3.us4.outblaze.com [205.158.62.25]) by sdf.lonestar.org (8.12.10/8.12.10) with ESMTP id iAELE1jo011913 for ; Sun, 14 Nov 2004 21:14:04 GMT Received: from lists.gnu.org (lists.gnu.org [199.232.76.165]) by spf3.us4.outblaze.com (Postfix) with ESMTP id A885354912 for ; Sun, 14 Nov 2004 21:13:34 +0000 (GMT) Received: from localhost ([127.0.0.1] helo=lists.gnu.org) by lists.gnu.org with esmtp (Exim 4.33) id 1CTRpQ-0003x8-4m for migo@homemail.com; Sun, 14 Nov 2004 16:22:32 -0500 Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.33) id 1CTRoK-0003ho-K5 for gnu-arch-users@gnu.org; Sun, 14 Nov 2004 16:21:24 -0500 Received: from exim by lists.gnu.org with spam-scanned (Exim 4.33) id 1CTRoI-0003gw-58 for gnu-arch-users@gnu.org; Sun, 14 Nov 2004 16:21:22 -0500 Received: from [199.232.76.173] (helo=monty-python.gnu.org) by lists.gnu.org with esmtp (Exim 4.33) id 1CTRoH-0003gi-Vd for gnu-arch-users@gnu.org; Sun, 14 Nov 2004 16:21:21 -0500 Received: from [216.254.0.203] (helo=mail3.speakeasy.net) by monty-python.gnu.org with esmtp (TLSv1:DES-CBC3-SHA:168) (Exim 4.34) id 1CTRfM-0001rX-KQ for gnu-arch-users@gnu.org; Sun, 14 Nov 2004 16:12:08 -0500 Received: (qmail 32533 invoked from network); 14 Nov 2004 21:12:07 -0000 Received: from dsl093-114-095.chi2.dsl.speakeasy.net (HELO mofo.meme.com) (kop@[66.93.114.95]) (envelope-sender ) by mail3.speakeasy.net (qmail-ldap-1.03) with SMTP for ; 14 Nov 2004 21:12:07 -0000 Received: from mofo.meme.com (localhost [127.0.0.1]) by mofo.meme.com (Postfix) with ESMTP id E0AA54361; Sun, 14 Nov 2004 15:23:29 -0600 (CST) Date: Sun, 14 Nov 2004 15:23:29 -0600 From: "Karl O. Pinc" To: John Meinel Subject: Re: [Gnu-arch-users] Removing the last changeset(s) from the archive Message-ID: <20041114152329.C29200@mofo.meme.com> References: <20041114111431.E15533@mofo.meme.com> <4197AAA5.7080501@panoramicfeedback.com> <20041114131902.Q15533@mofo.meme.com> <1100459623.15201.46.camel@localhost> <20041114134530.A23046@mofo.meme.com> <4197BF40.1010300@arbash-meinel.com> Mime-Version: 1.0 Content-Type: text/plain; format=flowed; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In-Reply-To: <4197BF40.1010300@arbash-meinel.com>; from john@arbash-meinel.com on Sun, Nov 14, 2004 at 14:25:36 -0600 X-Mailer: Balsa 1.2.4 Cc: gnu-arch-users@gnu.org, Charles Duffy X-BeenThere: gnu-arch-users@gnu.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: a discussion list for all things arch-ish List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: gnu-arch-users-bounces+migo=homemail.com@gnu.org Errors-To: gnu-arch-users-bounces+migo=homemail.com@gnu.org Status: RO Content-Length: 2309 Lines: 52 On 2004.11.14 14:25 John Meinel wrote: > I think you've just described the difference between a dev tree, and > a stable tree. > But if you really want "this tree is always good", you probably want > something more like a stable branch, rather than trying to do that > with your dev branch. When I'm developing I'm partly using revision control to keep track of what I'm working on. I just don't want it to get in the way, so, especially while I'm learning arch I don't want to have to stop and figure out how to undo a commit mistake. But that's niether here nor there because the'll always be mistakes. In terms of using revision control as a brain extension, I'd like to be able to try things out in my development branch with the freedom allowed when I always know I can undo, so I want the commits in the archive to be "clean" cuts of what I think I've done. As I often am doing more than one thing at a time this means committing with care, to ensure that 'one thing' gets committed at a time. Certainly there's always a fixup when the rule is "archive changes are permanent", but there's more brainwork when recovering from a commit mistake, and more brain work should one of my current ideas dead-end and I have to make sense of what to revert. (I'm not talking here of un-committing a recent change but of backing out something old and then commiting that revert to the tree.) A quick 'undo-commit' after making a commit mistake would be handy, but not essential. Anybody sharing an archive would have to live without 'undo-commit', given that arch has no network protocol. In some sense sharing can be construed as the difference between a stable and a deveopment branch. Because many will have to live without 'undo-commit' I wouldn't think it worth doing unless it can be done easily and cleanly. Absent an undo-commit, a 'standardized' interface for replay -reverse;sync-tree;commit would reduce the "Oh no!" factor of making a commit mistake. Karl Free Software: "You don't pay back, you pay forward." -- Robert A. Heinlein _______________________________________________ Gnu-arch-users mailing list Gnu-arch-users@gnu.org http://lists.gnu.org/mailman/listinfo/gnu-arch-users GNU arch home page: http://savannah.gnu.org/projects/gnu-arch/