From mboxrd@z Thu Jan 1 00:00:00 1970 From: =?utf-8?Q?Ludovic_Court=C3=A8s?= Subject: Re: User shell: state or config? Date: Sat, 27 Apr 2019 12:51:03 +0200 Message-ID: <87pnp7iuw8.fsf@gnu.org> References: <874l6mpduo.fsf@gnu.org> <87sgu60xor.fsf@nckx> <87r29pb7vz.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]:39480) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1hKKvG-0004HO-4K for guix-devel@gnu.org; Sat, 27 Apr 2019 06:51:11 -0400 In-Reply-To: <87r29pb7vz.fsf@gmail.com> (Chris Marusich's message of "Thu, 25 Apr 2019 23:25:20 -0700") 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: Chris Marusich Cc: guix-devel@gnu.org Hello, Chris Marusich skribis: > Tobias Geerinckx-Rice writes: > >>> All in all, I=E2=80=99m in favor of switching back to the previous beha= vior: >>> considering user shells as system config. That=E2=80=99s a one-line ch= ange >>> in >>> (gnu build accounts). >> >> +1! > > +1 from me as well. Done in 504a0fc636ec591e65b4a229a37e522e425d8a0c. > Allowing the shell to be declaratively defined in the system config, > but also allowing a user to override it permanently at runtime via > chsh, seems like a nice middle ground between treating it as "config" > vs. "state". In this particular case there=E2=80=99s no middle ground though: either you consider state and then you have the problems I listed, or you consider it config. I did think there was a middle ground, until I realized the implications. Thanks everyone for your feedback! Ludo=E2=80=99.