From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mark H Weaver Subject: bug#31841: ./pre-inst-env guix system no longer works without ~/.config/guix Date: Fri, 15 Jun 2018 09:08:00 -0400 Message-ID: <87tvq4nqof.fsf@netris.org> References: <8736xopz0q.fsf@netris.org> <87bmccuwbu.fsf@mdc-berlin.de> <87o9gctgrm.fsf@gnu.org> <8736xouuh7.fsf@mdc-berlin.de> 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]:41242) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1fToUQ-0001r3-Lu for bug-guix@gnu.org; Fri, 15 Jun 2018 09:10:07 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1fToUM-0001mi-MV for bug-guix@gnu.org; Fri, 15 Jun 2018 09:10:06 -0400 Received: from debbugs.gnu.org ([208.118.235.43]:41197) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_128_CBC_SHA1:16) (Exim 4.71) (envelope-from ) id 1fToUM-0001ma-Hq for bug-guix@gnu.org; Fri, 15 Jun 2018 09:10:02 -0400 Received: from Debian-debbugs by debbugs.gnu.org with local (Exim 4.84_2) (envelope-from ) id 1fToUM-0006bH-5z for bug-guix@gnu.org; Fri, 15 Jun 2018 09:10:02 -0400 Sender: "Debbugs-submit" Resent-Message-ID: In-Reply-To: <8736xouuh7.fsf@mdc-berlin.de> (Ricardo Wurmus's message of "Fri, 15 Jun 2018 14:04:04 +0200") List-Id: Bug reports for GNU Guix List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: bug-guix-bounces+gcggb-bug-guix=m.gmane.org@gnu.org Sender: "bug-Guix" To: Ricardo Wurmus Cc: 31841@debbugs.gnu.org Hi, Ricardo Wurmus writes: > Ludovic Court=C3=A8s writes: > >> Hello, >> >> Ricardo Wurmus skribis: >> >>> When outside of an environment containing guile-sqlite3 strace shows >>> that Guix tries to load it right before failing. The cause here is the >>> autoloading of (guix store database), which depends on the sqlite >>> bindings. >> >> Indeed. Commit c5a2e1ffcb029f50c4c18352cf378b61c41c625e made >> guile-sqlite3 a requirement, but =E2=80=98configure=E2=80=99 is supposed= to error out if >> it=E2=80=99s not found. >> >> Mark, any idea why =E2=80=98configure=E2=80=99 didn=E2=80=99t fail for y= ou? What does >> =E2=80=98config.log=E2=80=99 say? > > I have the same problem when using =E2=80=9Cpre-inst-env=E2=80=9D outside= of a > development environment. So =E2=80=9Cconfigure=E2=80=9D succeeds (when r= un inside =E2=80=9Cguix > enviromnent guix=E2=80=9D), but =E2=80=9C./pre-inst-env guix system --hel= p=E2=80=9D fails when > run outside of that environment. Indeed. I make sure to run 'make' within a development environment, but until now I've been able to *run* Guix using ./pre-inst-env outside of the development environment. My preferred mode of operation has always been to run 'guix' from my private git checkout. For years I was able to arrange this by manually installing a symlink from ~/.config/guix/latest to my git checkout, but at some point that stopped working, so now I have a small 'guix' script in ~mhw/bin and ~root/bin: #!/bin/sh exec /home/mhw/guix/pre-inst-env guix "$@" I suppose I could extend this script to run 'pre-inst-env' within the development environment, but I wonder if there's a reason not to have 'pre-inst-env' install the necessary environment. That would seem to be its job, no? Mark