From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeff King Subject: Re: [PATCH 0/2] merge --summary vs. merge.summary Date: Thu, 3 Apr 2008 06:30:56 -0400 Message-ID: <20080403103056.GD6673@coredump.intra.peff.net> References: <1207213396-17370-1-git-send-email-szeder@ira.uka.de> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: git@vger.kernel.org To: SZEDER =?utf-8?B?R8OhYm9y?= X-From: git-owner@vger.kernel.org Thu Apr 03 12:31:47 2008 Return-path: Envelope-to: gcvg-git-2@gmane.org Received: from vger.kernel.org ([209.132.176.167]) by lo.gmane.org with esmtp (Exim 4.50) id 1JhMjP-0006QE-Vx for gcvg-git-2@gmane.org; Thu, 03 Apr 2008 12:31:44 +0200 Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753275AbYDCKbA convert rfc822-to-quoted-printable (ORCPT ); Thu, 3 Apr 2008 06:31:00 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1753240AbYDCKa7 (ORCPT ); Thu, 3 Apr 2008 06:30:59 -0400 Received: from 66-23-211-5.clients.speedfactory.net ([66.23.211.5]:3856 "EHLO peff.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752806AbYDCKa7 (ORCPT ); Thu, 3 Apr 2008 06:30:59 -0400 Received: (qmail 10116 invoked by uid 111); 3 Apr 2008 10:30:57 -0000 Received: from coredump.intra.peff.net (HELO coredump.intra.peff.net) (10.0.0.2) by peff.net (qpsmtpd/0.32) with SMTP; Thu, 03 Apr 2008 06:30:57 -0400 Received: by coredump.intra.peff.net (sSMTP sendmail emulation); Thu, 03 Apr 2008 06:30:56 -0400 Content-Disposition: inline In-Reply-To: <1207213396-17370-1-git-send-email-szeder@ira.uka.de> Sender: git-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: git@vger.kernel.org Archived-At: On Thu, Apr 03, 2008 at 11:03:14AM +0200, SZEDER G=C3=A1bor wrote: > So the '--summary' option controls the printing of diffstat, while th= e > 'merge.summary' config variable controls the behaviour of 'fmt-merge-= msg', Think that --diffstat is probably a better name. But your series actually _changes_ the meaning of --summary, which is probably a bad idea for people who have scripted around these. The problem is that you have two concepts with the same name; giving on= e of the concepts the name means that the other concept breaks. I think the most compatible way to go forward is to use two new names: 1. --summary (and --no- form) changes to --diffstat; --summary remain= s as a compatibility alias for --diffstat. 2. merge.summary becomes merge.something_else, along with --something-else support. merge.summary remains as a compatibility option for merge.something_else. Though I can't think of a good "something_else". Given that there isn't even a command-line equivalent to merge.summary now, maybe just leave i= t as merge.summary, stop advertising --summary (even though it is a compatibility alias, we officially call it --diffstat), and then there is less confusion on the part of documentation readers (though perhaps more on the part of those who blindly try --summary). -Peff