From mboxrd@z Thu Jan 1 00:00:00 1970 From: "David S. Miller" Subject: Re: non-trivial merge failures Date: Sun, 20 Nov 2005 20:20:32 -0800 (PST) Message-ID: <20051120.202032.37639519.davem@davemloft.net> References: <20051120.134945.104623647.davem@davemloft.net> <20051120230738.GA16861@mythryan2.michonline.com> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: git@vger.kernel.org X-From: git-owner@vger.kernel.org Mon Nov 21 05:21:38 2005 Return-path: Received: from vger.kernel.org ([209.132.176.167]) by ciao.gmane.org with esmtp (Exim 4.43) id 1Ee3AB-0002dR-Pi for gcvg-git@gmane.org; Mon, 21 Nov 2005 05:20:20 +0100 Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1750970AbVKUEUR (ORCPT ); Sun, 20 Nov 2005 23:20:17 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932145AbVKUEUQ (ORCPT ); Sun, 20 Nov 2005 23:20:16 -0500 Received: from dsl027-180-168.sfo1.dsl.speakeasy.net ([216.27.180.168]:38573 "EHLO sunset.davemloft.net") by vger.kernel.org with ESMTP id S1750970AbVKUEUP (ORCPT ); Sun, 20 Nov 2005 23:20:15 -0500 Received: from localhost ([127.0.0.1] ident=davem) by sunset.davemloft.net with esmtp (Exim 4.54) id 1Ee3AP-0005zs-1w; Sun, 20 Nov 2005 20:20:33 -0800 To: ryan@michonline.com In-Reply-To: <20051120230738.GA16861@mythryan2.michonline.com> X-Mailer: Mew version 4.2.53 on Emacs 21.4 / Mule 5.0 (SAKAKI) Sender: git-owner@vger.kernel.org Precedence: bulk X-Mailing-List: git@vger.kernel.org Archived-At: From: Ryan Anderson Date: Sun, 20 Nov 2005 18:07:38 -0500 > On Sun, Nov 20, 2005 at 01:49:45PM -0800, David S. Miller wrote: > > > > Every time I try to do a non-trivial merge, I always > > get this: > > git-merge-recursive became the default merge tool recently, and Debian > uses Python 2.3 as the default, which lacks "subprocess.py". Luckily, > git ships a backport, so you can do: > > make WITH_OWN_SUBPROCESS_PY=YesPlease all doc test install install-doc > or > make PYTHON_PATH=/usr/bin/python2.4 all doc test install install-doc All these ways to make it work are nice, and in fact I have no problem with upgrading to python2.4 on my system if that makes it work. I have a huge problem with the fact that "make test" passes %100 cleanly even if you have the wrong python version installed. It's amazing that this is that possible if even the most non-trivial merge needs this Python code :-)