From mboxrd@z Thu Jan 1 00:00:00 1970 From: =?utf-8?Q?Ludovic_Court=C3=A8s?= Subject: Re: Merging =?utf-8?Q?=E2=80=98wip-newt-installer=E2=80=99?= in master? Date: Wed, 16 Jan 2019 23:28:29 +0100 Message-ID: <87bm4gteya.fsf@gnu.org> References: <87lg6sejwa.fsf@gmail.com> <877ehbydmz.fsf@gmail.com> <87sgzykmyx.fsf@gnu.org> <877eh9245u.fsf@gmail.com> <87efbgx0p0.fsf@gnu.org> <87sgzw5ylb.fsf@gmail.com> <87zhu1cw4p.fsf_-_@gnu.org> <875zwp17od.fsf@gmail.com> <87lg5j6h00.fsf@gnu.org> <87pnuv26jv.fsf@gmail.com> <87efbb4xpt.fsf@gnu.org> <871s7bm8k5.fsf@gmail.com> <8736rl6049.fsf@gmail.com> <874lc15opg.fsf@gnu.org> <87r2f4smlk.fsf@gmail.com> <8736rk6ui1.fsf@gnu.org> <87d0qgt8ej.fsf@gmail.com> <871s5q1zyv.fsf@gnu.org> <874ladwudb.fsf@gmail.com> <874la8v84p.fsf@gnu.org> <87won4v658.fsf@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Return-path: Received: from eggs.gnu.org ([209.51.188.92]:44156) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1gjtfn-0000j6-7p for guix-devel@gnu.org; Wed, 16 Jan 2019 17:28:35 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1gjtfm-0006EU-6Q for guix-devel@gnu.org; Wed, 16 Jan 2019 17:28:35 -0500 Received: from hera.aquilenet.fr ([185.233.100.1]:57176) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1gjtfl-00068Y-TQ for guix-devel@gnu.org; Wed, 16 Jan 2019 17:28:34 -0500 In-Reply-To: <87won4v658.fsf@gmail.com> (Mathieu Othacehe's message of "Wed, 16 Jan 2019 18:55:47 +0100") List-Id: "Development of GNU Guix and the GNU System distribution." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: guix-devel-bounces+gcggd-guix-devel=m.gmane.org@gnu.org Sender: "Guix-devel" To: Mathieu Othacehe Cc: Guix-devel Hi! Mathieu Othacehe skribis: > Sadly I think its kind of the same problem as the one I described here: > https://lists.gnu.org/archive/html/guix-devel/2018-12/msg00161.html. > > As we are not able to undo what's done by the cow-store service, > once restarted, the installer can't be used properly anymore. True, that=E2=80=99s indeed what you described before. > So I guess the solution to both of these problems would be to force a > reboot if the installer fails after the cow-store service has been > started. Hmm yes, maybe? Or maybe we can unshare(2) the installer process so it still sees the original file system layout? Or, conversely, run the installation process in a separate name space? Needs more thought=E2=80=A6 Ludo=E2=80=99.