From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeff King Subject: Re: BUG:git-log --follow should never trim output WRT equivalent git-log Date: Mon, 24 Mar 2008 21:45:04 -0400 Message-ID: <20080325014504.GA15864@coredump.intra.peff.net> References: Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Cc: git@vger.kernel.org To: Santiago Gala X-From: git-owner@vger.kernel.org Tue Mar 25 02:45:53 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 1JdyEX-0007Cu-AP for gcvg-git-2@gmane.org; Tue, 25 Mar 2008 02:45:49 +0100 Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753994AbYCYBpI (ORCPT ); Mon, 24 Mar 2008 21:45:08 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754022AbYCYBpI (ORCPT ); Mon, 24 Mar 2008 21:45:08 -0400 Received: from 66-23-211-5.clients.speedfactory.net ([66.23.211.5]:3530 "EHLO peff.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753552AbYCYBpH (ORCPT ); Mon, 24 Mar 2008 21:45:07 -0400 Received: (qmail 32246 invoked by uid 111); 25 Mar 2008 01:45:06 -0000 Received: from coredump.intra.peff.net (HELO coredump.intra.peff.net) (10.0.0.2) by peff.net (qpsmtpd/0.32) with SMTP; Mon, 24 Mar 2008 21:45:06 -0400 Received: by coredump.intra.peff.net (sSMTP sendmail emulation); Mon, 24 Mar 2008 21:45:04 -0400 Content-Disposition: inline In-Reply-To: Sender: git-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: git@vger.kernel.org Archived-At: On Mon, Mar 24, 2008 at 04:19:01PM +0000, Santiago Gala wrote: > $ git log -M -C -C --color --stat --follow -- gitk-git/po/de.po | wc -l > 0 > sgala@marlow ~/newcode/git/git (master)$ git log -M -C -C --color --stat -- > gitk-git/po/de.po | wc -l > 31 > $ git --version > git version 1.5.4.4 > > I find very strange that --follow makes git log to stop outputting the first > entries. I think the tricky thing here is that every one of those commits is a merge, which doesn't actually "introduce" any content, and that probably confuses the "--follow" code. The thing is that getting the history of gitk-git specifically is a bit tricky. All of the commits actually happen to, e.g., the file 'gitk' in the main gitk repository, and then Junio merges it into git using a special "subtree" merge as 'gitk-git/gitk'. So looking at a path in 'gitk-git' will always just turn up merge commits; the actual commits took place on a different path. But I haven't looked too closely at what's going on, so I may be way off base. My two cents. -Peff