From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jason Riedy Subject: Re: [PATCH] Pass --upload-pack and --receive-pack through submodules. Date: Thu, 01 Jan 2009 20:44:03 -0500 Message-ID: <878wpua3do.fsf@sparse.dyndns.org> References: <1230605055-30723-1-git-send-email-jason@acm.org> <1230649782-14710-1-git-send-email-jason@acm.org> <7vy6xuhhbx.fsf@gitster.siamese.dyndns.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: git@vger.kernel.org To: Junio C Hamano X-From: git-owner@vger.kernel.org Fri Jan 02 02:45:47 2009 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 1LIZ6h-0004wG-8H for gcvg-git-2@gmane.org; Fri, 02 Jan 2009 02:45:47 +0100 Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755449AbZABBoZ (ORCPT ); Thu, 1 Jan 2009 20:44:25 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751702AbZABBoZ (ORCPT ); Thu, 1 Jan 2009 20:44:25 -0500 Received: from a.mail.sonic.net ([64.142.16.245]:54070 "EHLO a.mail.sonic.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750911AbZABBoY (ORCPT ); Thu, 1 Jan 2009 20:44:24 -0500 Received: from nan.sparse.yi.org (misc-148-78-88-138.pool.starband.net [148.78.88.138]) (authenticated bits=0) by a.mail.sonic.net (8.13.8.Beta0-Sonic/8.13.7) with ESMTP id n021iAEw018348; Thu, 1 Jan 2009 17:44:13 -0800 In-Reply-To: <7vy6xuhhbx.fsf@gitster.siamese.dyndns.org> (Junio C. Hamano's message of "Thu, 01 Jan 2009 13:00:50 -0800") User-Agent: Gnus/5.110011 (No Gnus v0.11) Emacs/23.0.60 (gnu/linux) Sender: git-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: git@vger.kernel.org Archived-At: And Junio C. Hamano writes: > Do we (and can we) assume that the remote repositories submodules fetch > from all reside on the same host and can share the same values for these > parameters? Shouldn't these instead be specified in the configuration > files for the submodules, if they need to be nonstandard values? git does not and (for me) should not make the assumption above. I'm fetching different pieces from different administrative domains where I also have push access. While I could clone them all into one place, that adds another level of repositories I'd have to keep up-to-date. Essentially, I'm collecting experimental libraries and drivers into one build tree. My intent is to record exactly what is being built on multiple machines while being able to push local build configurations back to my central superproject. Then I can make a version with pull-only remote URLs[1] for others to duplicate the setup for my experiments. I suppose I could clone the submodules by hand and then use git submodule add. But then I'll have to build the structure by hand (with my own wrapper) everywhere I use the supermodule. At that point, using git submodule becomes a tad silly. Or I could set up the superproject with pull-only URLs and modify the configurations with a separate wrapper. Seemed like adding the two options to git submodule was easier. Jason Footnotes: [1] Or ask for separate pull- and push-urls in git remote. But I haven't thought that through.