From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Shawn O. Pearce" Subject: Re: What's cooking in git.git (Aug 2008, #01; Tue, 05) Date: Wed, 6 Aug 2008 07:21:28 -0700 Message-ID: <20080806142128.GA28231@spearce.org> References: <7vmyjqsm04.fsf@gitster.siamese.dyndns.org> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Cc: git@vger.kernel.org To: Junio C Hamano X-From: git-owner@vger.kernel.org Wed Aug 06 16:22:44 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 1KQjuJ-0003Ww-MM for gcvg-git-2@gmane.org; Wed, 06 Aug 2008 16:22:32 +0200 Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756060AbYHFOV3 (ORCPT ); Wed, 6 Aug 2008 10:21:29 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754908AbYHFOV3 (ORCPT ); Wed, 6 Aug 2008 10:21:29 -0400 Received: from george.spearce.org ([209.20.77.23]:51678 "EHLO george.spearce.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752612AbYHFOV2 (ORCPT ); Wed, 6 Aug 2008 10:21:28 -0400 Received: by george.spearce.org (Postfix, from userid 1001) id 4141438420; Wed, 6 Aug 2008 14:21:28 +0000 (UTC) Content-Disposition: inline In-Reply-To: <7vmyjqsm04.fsf@gitster.siamese.dyndns.org> User-Agent: Mutt/1.5.17+20080114 (2008-01-14) Sender: git-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: git@vger.kernel.org Archived-At: Junio C Hamano wrote: > [New Topics] > * sp/smart-http (Sun Aug 3 00:25:17 2008 -0700) 2 commits > - [do not merge -- original version] Add Git-aware CGI for Git-aware > smart HTTP transport > - Add backdoor options to receive-pack for use in Git-aware CGI > > The "magic" detection protocol was revised to use POST to info/refs; the > top one queued is from before that discussion. I'm surprised you queued these. They really are not ready for application to a tree. For example the CGI is really messed up for trying to do Transfer-Encoding: chunked on its own, JH was right and Apache does it for us. That removed a good chunk of code from it. Writing the client for this is slightly non-trivial. I more-or-less want to shove it into send-pack, not http-push, as the bulk of the protocol we need to speak is there in send-pack. Anyway, I'm going to try to get back to this topic more later this week. I have been focusing on JGit this week (or at least trying to) as I really need to get some stuff done there for EGit. -- Shawn.