From mboxrd@z Thu Jan  1 00:00:00 1970
From: ludo@gnu.org (Ludovic =?utf-8?Q?Court=C3=A8s?=)
Subject: Topic branches
Date: Sun, 01 Sep 2013 22:13:38 +0200
Message-ID: <878uzgfggd.fsf_-_@gnu.org>
References: <1377644626-17292-1-git-send-email-tipecaml@gmail.com>
	<8761ukoct3.fsf@gnu.org> <20130901144110.GA23502@debian>
	<874na4ib4k.fsf@gnu.org> <20130901195046.GA4912@debian>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Return-path: <guix-devel-bounces+gcggd-guix-devel=m.gmane.org@gnu.org>
Received: from eggs.gnu.org ([2001:4830:134:3::10]:37263)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <ludo@gnu.org>) id 1VGE6h-0002CF-ED
	for guix-devel@gnu.org; Sun, 01 Sep 2013 16:18:53 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <ludo@gnu.org>) id 1VGE6b-00037R-6t
	for guix-devel@gnu.org; Sun, 01 Sep 2013 16:18:47 -0400
Received: from hera.aquilenet.fr ([141.255.128.1]:59470)
	by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from <ludo@gnu.org>)
	id 1VGE6b-00037N-0S
	for guix-devel@gnu.org; Sun, 01 Sep 2013 16:18:41 -0400
In-Reply-To: <20130901195046.GA4912@debian> (Andreas Enge's message of "Sun, 1
	Sep 2013 21:50:46 +0200")
List-Id: <guix-devel.gnu.org>
List-Unsubscribe: <https://lists.gnu.org/mailman/options/guix-devel>,
	<mailto:guix-devel-request@gnu.org?subject=unsubscribe>
List-Archive: <http://lists.gnu.org/archive/html/guix-devel>
List-Post: <mailto:guix-devel@gnu.org>
List-Help: <mailto:guix-devel-request@gnu.org?subject=help>
List-Subscribe: <https://lists.gnu.org/mailman/listinfo/guix-devel>,
	<mailto:guix-devel-request@gnu.org?subject=subscribe>
Errors-To: guix-devel-bounces+gcggd-guix-devel=m.gmane.org@gnu.org
Sender: guix-devel-bounces+gcggd-guix-devel=m.gmane.org@gnu.org
To: Andreas Enge <andreas@enge.fr>
Cc: guix-devel@gnu.org

Andreas Enge <andreas@enge.fr> skribis:

> On Sun, Sep 01, 2013 at 09:40:27PM +0200, Ludovic Court=C3=A8s wrote:
>> Speaking of which: in the future, we should use topic branches for such
>> things, to avoid disrupting the main branch.  For instance, I=E2=80=99ve=
 locally
>> switched back to Python 2 as the default since I was otherwise unable to
>> use the QEMU-related things.
>>=20
>> Actually it=E2=80=99s not too late: we could create a new branch off =E2=
=80=98master=E2=80=99,
>> and just switch back to Python 2 as the default on =E2=80=98master=E2=80=
=99.  Hydra can
>> be told to build the new branch in addition to =E2=80=98master=E2=80=99.
>
> Actually, 77c7f8f41b558bab13690c843068af8ba996e5bf switches back (while
> keeping the definition of Python 3 in the variable python-3; but all
> packages using python as input will get Python 2).

Ah right, thanks!

> We could create branches; very often (like here, with changes to the pyth=
on
> build system pending), we might as well use core-updates directly.

Experience with Nixpkgs suggests that topic branches are better:
core-updates does rebuild-the-world changes, so you don=E2=80=99t want to m=
erge
it lightly, and not too often; conversely the Python changes are
relatively isolated.

More importantly, intermingling the effects of the Python changes with
that of (say) a GCC and glibc upgrade is undesirable, as it makes it
harder to determine the cause of a problem.

Thanks,
Ludo=E2=80=99.