From mboxrd@z Thu Jan 1 00:00:00 1970 From: Magnus Hjorth Subject: Re: Committing with past date? Date: Tue, 02 Sep 2008 04:42:14 +0200 Message-ID: <1220323334.5476.29.camel@magnus-desktop> References: <1220179469.5518.5.camel@magnus-desktop> <1220286321.5474.6.camel@magnus-desktop> <200809020231.32780.jnareb@gmail.com> Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit Cc: git@vger.kernel.org To: Jakub Narebski X-From: git-owner@vger.kernel.org Tue Sep 02 04:43:56 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 1KaLs3-0000sL-HK for gcvg-git-2@gmane.org; Tue, 02 Sep 2008 04:43:56 +0200 Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752792AbYIBCmU (ORCPT ); Mon, 1 Sep 2008 22:42:20 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752516AbYIBCmU (ORCPT ); Mon, 1 Sep 2008 22:42:20 -0400 Received: from ch-smtp02.sth.basefarm.net ([80.76.149.213]:37563 "EHLO ch-smtp02.sth.basefarm.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752557AbYIBCmT (ORCPT ); Mon, 1 Sep 2008 22:42:19 -0400 Received: from c83-254-118-193.bredband.comhem.se ([83.254.118.193]:52017) by ch-smtp02.sth.basefarm.net with esmtp (Exim 4.68) (envelope-from ) id 1KaLqU-0001l9-6v; Tue, 02 Sep 2008 04:42:18 +0200 In-Reply-To: <200809020231.32780.jnareb@gmail.com> X-Mailer: Evolution 2.22.3.1 X-Originating-IP: 83.254.118.193 X-Scan-Result: No virus found in message 1KaLqU-0001l9-6v. X-Scan-Signature: ch-smtp02.sth.basefarm.net 1KaLqU-0001l9-6v ef62f978f081eef3c6afc4f4eb529807 Sender: git-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: git@vger.kernel.org Archived-At: On Tue, 2008-09-02 at 02:31 +0200, Jakub Narebski wrote: > On Mon, 1 Sep 2008, Magnus Hjorth wrote: > > On Sun, 2008-08-31 at 04:12 -0700, Jakub Narebski wrote: > > > Magnus Hjorth writes: > > > > > > > > > > I'm trying to port over old version history that I maintained manually > > > > (tarballs and changelogs) into a git repository. > > > > > > For that, I think it would be best to take a look at example > > > fast-import script: contrib/fast-import/import-tars.perl; > > > there is equivalent contrib/fast-import/import-zips.py if you > > > perfer either Pyhon over Perl, and/or zips over tarballs. > By the way, is there any reason _not_ to use import-tars.perl from > the contrib/fast-import in your case? Well, I was planning do a lot of manual cleanup of the files also (splitting commits, removing auto-generated files etc), so importing the tarballs would just be the first step. But I'll try it, it can still be useful to get a starting point... > > > > I have successfully imported the CVS history using git-cvsimport, but > > now I want to add these older revisions that were made with tarballs to > > the same tree, before the CVS revisions. The last tarball and the first > > CVS revision have identical content, and I would like to somehow "glue" > > the histories together. > > > > Can this be done? > > It can be done for example using grafts. Search git mailing list for > graftshistory (or something like that) script, which was used to join > using grafts git "current work" Linux repository (started from "scratch"). > Then you can check in gitk if everything is all right. If you truly > require connected histories, and not being able to locally turn on > and off the historical repository, you can always use git-filter-branch > which (among others) can turn grafts into true commits. The grafts approach looks interesting. I still need some way to create multiple unconnected trees in one repo. Maybe the safest way to do that is to create a separate historic repo first and then transfer the historic branch using git-bundle to my main working repo. Regards, Magnus