From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Kastrup Subject: Re: git gc --aggressive led to about 40 times slower "git log --raw" Date: Wed, 19 Feb 2014 10:01:08 +0100 Message-ID: <87bny3qxwr.fsf@fencepost.gnu.org> References: <87r470ssuc.fsf@fencepost.gnu.org> <87ioscsoow.fsf@fencepost.gnu.org> <20140218155842.GA7855@google.com> Mime-Version: 1.0 Content-Type: text/plain Cc: Duy Nguyen , Junio C Hamano , Jonathan Nieder , Christian Jaeger , Git Mailing List To: Philippe Vaucher X-From: git-owner@vger.kernel.org Wed Feb 19 10:01:21 2014 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 1WG31r-0001OE-0c for gcvg-git-2@plane.gmane.org; Wed, 19 Feb 2014 10:01:19 +0100 Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752122AbaBSJBL (ORCPT ); Wed, 19 Feb 2014 04:01:11 -0500 Received: from fencepost.gnu.org ([208.118.235.10]:35559 "EHLO fencepost.gnu.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751375AbaBSJBJ (ORCPT ); Wed, 19 Feb 2014 04:01:09 -0500 Received: from localhost ([127.0.0.1]:34600 helo=lola) by fencepost.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1WG31g-0002WJ-Hh; Wed, 19 Feb 2014 04:01:08 -0500 Received: by lola (Postfix, from userid 1000) id 08141E0683; Wed, 19 Feb 2014 10:01:08 +0100 (CET) In-Reply-To: (Philippe Vaucher's message of "Wed, 19 Feb 2014 09:38:31 +0100") User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/24.3.50 (gnu/linux) Sender: git-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: git@vger.kernel.org Archived-At: Philippe Vaucher writes: >> fwiw this is the thread that added --depth=250 >> >> http://thread.gmane.org/gmane.comp.gcc.devel/94565/focus=94626 > > This post is quite interesting: > http://article.gmane.org/gmane.comp.gcc.devel/94637 Yes. Of course I am prejudiced because I volunteered fixing git-blame on the Emacs developer list in order to make it more feasible to transfer the Emacs repository to Git. Calling git blame via C-x v g is a rather important part of the workflow, and it's currently intolerable to work with on a number of files. While I'm fixing the basic shortcomings in builtin/blame.c itself, the operation "fetch the objects" is necessary for all objects at least once. It's conceivable that some nice caching strategy would help with avoiding the repeated traversal of long delta chain tails. That could also help defusing the operation of basic stuff like git-log. But the short and long end of it is that there are valid operations accessing a large amount of past history, and one point of having a distributed version control system with non-shallow repository by default is to have history and ways of working with it at one's hand. And git's default modus of operation is _not_ to store things like copies and moves and renames in commits, but deduce them from looking at the stored data. So making looking at stored data including old data expensive means that Git does not work well in the way it is designed to operate. -- David Kastrup