From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Wong Subject: Re: git-svn is "Unable to determine upstream SVN information..." Date: Fri, 10 Jul 2009 13:33:22 -0700 Message-ID: <20090710203322.GA18051@dcvr.yhbt.net> References: <20090710033133.GA23082@dcvr.yhbt.net> <5m5e55538tlnq40rup7rci61i2mg16q0dv@4ax.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: git@vger.kernel.org To: Esben Skovenborg X-From: git-owner@vger.kernel.org Fri Jul 10 22:33:49 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 1MPMmy-0006h0-BZ for gcvg-git-2@gmane.org; Fri, 10 Jul 2009 22:33:48 +0200 Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753916AbZGJUdY (ORCPT ); Fri, 10 Jul 2009 16:33:24 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752302AbZGJUdX (ORCPT ); Fri, 10 Jul 2009 16:33:23 -0400 Received: from dcvr.yhbt.net ([64.71.152.64]:59697 "EHLO dcvr.yhbt.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751401AbZGJUdX (ORCPT ); Fri, 10 Jul 2009 16:33:23 -0400 Received: from localhost (unknown [127.0.2.5]) by dcvr.yhbt.net (Postfix) with ESMTP id D20D51F5FC; Fri, 10 Jul 2009 20:33:22 +0000 (UTC) Content-Disposition: inline In-Reply-To: <5m5e55538tlnq40rup7rci61i2mg16q0dv@4ax.com> User-Agent: Mutt/1.5.18 (2008-05-17) Sender: git-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: git@vger.kernel.org Archived-At: Esben Skovenborg wrote: > On Thu, 9 Jul 2009 20:31:33 -0700, Eric Wong wrote: > >Are there any weird characters in the URL? > > No. The only thing 'different' about this svn server is that it uses > authentication (i.e. asks for name+password). But I suppose that when the > git-svn fetch goes well, then this is not the problem? Do the git-svn > info and git-svn rebase even need to access the svn server? Yes, rebase (without the -l) option needs it for sure. I think info can optionally use it, too. Does enabling authentication caching (enabled by default in SVN) help things? I haven't used the authentication prompt pieces in ages, they're hard to test in an automated fashion so they may have bit-rotted since I last looked at them... > >Off the top of my head I can't think of anything else; I assume > >you're not allowed to share access to the repo (or to the clone) you're > >having problems with? > > Right, at least it's something we should talk about off-list. > > Is there any way of asking git-svn *why* it is "unable to determine > upstream SVN information from working tree history" ? If you're comfortable with a little Perl, you can sprinkle some debug statements in working_head_info(). Feel free to email me directly offlist but my schedule looks pretty hectic today and this weekend and I may be offline a lot. -- Eric Wong