From mboxrd@z Thu Jan 1 00:00:00 1970 From: Heiko Voigt Subject: Re: Re: Git issues with submodules Date: Fri, 29 Nov 2013 23:50:40 +0100 Message-ID: <20131129225040.GB31636@sandbox-ub> References: <20131122151120.GA32361@sigill.intra.peff.net> <528FC638.5060403@web.de> <20131122215454.GA4952@sandbox-ub> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Jens Lehmann , Ramkumar Ramachandra , Sergey Sharybin , Jeff King , Git List To: Junio C Hamano X-From: git-owner@vger.kernel.org Fri Nov 29 23:50:50 2013 Return-path: Envelope-to: gcvg-git-2@plane.gmane.org Received: from vger.kernel.org ([209.132.180.67]) by plane.gmane.org with esmtp (Exim 4.69) (envelope-from ) id 1VmWtb-0001pi-PV for gcvg-git-2@plane.gmane.org; Fri, 29 Nov 2013 23:50:48 +0100 Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752320Ab3K2Wuo (ORCPT ); Fri, 29 Nov 2013 17:50:44 -0500 Received: from smtprelay03.ispgateway.de ([80.67.31.26]:40630 "EHLO smtprelay03.ispgateway.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751068Ab3K2Wun (ORCPT ); Fri, 29 Nov 2013 17:50:43 -0500 Received: from [77.20.34.36] (helo=sandbox-ub) by smtprelay03.ispgateway.de with esmtpsa (TLSv1:AES128-SHA:128) (Exim 4.68) (envelope-from ) id 1VmWtW-0006hy-0i; Fri, 29 Nov 2013 23:50:42 +0100 Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.21 (2010-09-15) X-Df-Sender: aHZvaWd0QGh2b2lndC5uZXQ= Sender: git-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: git@vger.kernel.org Archived-At: On Mon, Nov 25, 2013 at 12:53:45PM -0800, Junio C Hamano wrote: > Heiko Voigt writes: > > > What I think needs fixing here first is that the ignore setting should not > > apply to any diffs between HEAD and index. IMO, it should only apply > > to the diff between worktree and index. > > Hmph. How about "git diff $commit", the diff between the worktree and > a named commit (which may or may not be HEAD)? Thats an interesting question. My first thought was that I would expect it to not show submodules since it involves the worktree, but then I could also argue that it should only show differences between whats in the index and the given commit. That would make matters more complicated but I image the use case (floating submodules) involves not caring about submodules except for some integration points when submodule sha1's are explicitly recorded. I would expect to only see diffs between these integration points. But then I am not a big user (none at all at the moment) of the floating model. Cheers Heiko