From mboxrd@z Thu Jan 1 00:00:00 1970 From: Yann Dirson Subject: Re: [PATCH 0/4] Re: Undocumented git-diff syntax Date: Mon, 11 Oct 2010 20:52:50 +0200 Message-ID: <20101011185250.GC4983@home.lan> References: <20101011150342.292ad725@chalon.bertin.fr> <20101011161721.35940919@chalon.bertin.fr> <20101011160224.GA25842@burratino> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Yann Dirson , Santi B?jar , git@vger.kernel.org To: Jonathan Nieder X-From: git-owner@vger.kernel.org Mon Oct 11 20:42:27 2010 Return-path: Envelope-to: gcvg-git-2@lo.gmane.org Received: from vger.kernel.org ([209.132.180.67]) by lo.gmane.org with esmtp (Exim 4.69) (envelope-from ) id 1P5NKM-0006ZF-OX for gcvg-git-2@lo.gmane.org; Mon, 11 Oct 2010 20:42:27 +0200 Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752051Ab0JKSmU (ORCPT ); Mon, 11 Oct 2010 14:42:20 -0400 Received: from smtp5-g21.free.fr ([212.27.42.5]:59844 "EHLO smtp5-g21.free.fr" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751769Ab0JKSmU (ORCPT ); Mon, 11 Oct 2010 14:42:20 -0400 Received: from home.lan (unknown [81.57.214.146]) by smtp5-g21.free.fr (Postfix) with ESMTP id 4940ED48217; Mon, 11 Oct 2010 20:42:12 +0200 (CEST) Received: from yann by home.lan with local (Exim 4.72) (envelope-from ) id 1P5NUQ-0003Fq-Qk; Mon, 11 Oct 2010 20:52:50 +0200 Content-Disposition: inline In-Reply-To: <20101011160224.GA25842@burratino> User-Agent: Mutt/1.5.20 (2009-06-14) Sender: git-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: git@vger.kernel.org Archived-At: On Mon, Oct 11, 2010 at 11:02:24AM -0500, Jonathan Nieder wrote: > Yann Dirson wrote: > > > OTOH, that appears not to be 100% accurate (my tests would have > > shown me): although --no-index does give me the expected results when > > one or two of the files are untracked, I have to explicitely add it, > > which seem to contradict this part of the doc. > > Yep, the doc is out of date. Luckily I had a series updating it > rotting in my local tree. > > Patch 5 gives --no-index its own prominent mention in the DESCRIPTION > section. I suspect that is a bad idea --- too much to learn for new > users --- and that the description should go under OPTIONS, but I ran > out of time. OTOH, not making it prominent does not help users to realize it is there. Keeping it last, or maybe in a "less frequently used" section looks fine to me - eg. the ... form could join it into such a section > Maybe patches 1-4 can help? They make things much more clear, thanks. So it looks like after this doc patch series, we're left with a couple of bugs: * the "fatal: '../Makefile' is outside repository" seen by Santi * the following which should probably complain that the files cannot be compared: $ echo a>a;echo b>b $ git diff a b; echo $? 0 At least, I can't think of them as meaningful behaviours. Best regards, -- Yann