From mboxrd@z Thu Jan 1 00:00:00 1970 From: George Clemmer Subject: Re: [doc RFC] Tame the Guix profile blizzard? Date: Tue, 10 Jul 2018 22:52:43 -0400 Message-ID: <87k1q2eb5w.fsf@gmail.com> References: <87tvp8q8pj.fsf@gmail.com> <87sh4qpwms.fsf@gnu.org> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Return-path: Received: from eggs.gnu.org ([2001:4830:134:3::10]:37406) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1fd5MS-0002qF-HU for guix-devel@gnu.org; Tue, 10 Jul 2018 23:00:13 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1fd5MP-0000q8-HQ for guix-devel@gnu.org; Tue, 10 Jul 2018 23:00:12 -0400 In-reply-to: <87sh4qpwms.fsf@gnu.org> 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: Ludovic =?utf-8?Q?Court=C3=A8s?= Cc: guix-devel , George Clemmer Ludovic Court=C3=A8s writes: > Hi George, > > George Clemmer skribis: > >> ISTM we can improve this situation as follows: >> >> 1) agree on a canonical term to use for the 4 types of profiles > > I think you mentioned only two types of =E2=80=9Cprofiles=E2=80=9D: one i= s profiles > created by =E2=80=98guix package=E2=80=99 or =E2=80=98guix environment=E2= =80=99 or =E2=80=98guix system=E2=80=99 (they > are all the same kind of =E2=80=9Cprofile=E2=80=9D), and the other one is= execution > profile, which has nothing to do with that. > > So to me it seems that =E2=80=9Cprofile=E2=80=9D is in fact used fairly c= onsistently, > isn=E2=80=99t it? Yes ... but ... the use of these profiles produces subtly different "side-effects". Package operations cause changes in the user's "default" profile that are immediately "used" in the user environment. However 'guix package -p foo' does not cause foo to be "used". 'guix environment --ad-hoc' causes the environment profile to be "added" to the user's environment, but --pure substitutes it for both system and "default" profiles IIUC. IMO these could be better explained and more easily referenced by users if we develop a structured nomenclature to distinquish between these. > However=E2=80=A6 > >> 3) add a top level discussion of profiles > > =E2=80=A6 we surely need this. Someone we should clearly define =E2=80= =9Cprofile=E2=80=9D > somewhere upfront. Perhaps we also need a glossary. Agreed