From mboxrd@z Thu Jan 1 00:00:00 1970 From: ludo@gnu.org (Ludovic =?utf-8?Q?Court=C3=A8s?=) Subject: Re: getting started with guix import nix Date: Mon, 19 Dec 2016 21:56:55 +0100 Message-ID: <8737hj3aeg.fsf@gnu.org> References: <1481632579.2545.11.camel@tourbillion-technology.com> <87oa0fjt92.fsf@gnu.org> <1481805189.2840.16.camel@tourbillion-technology.com> <87mvfs8d9l.fsf@gnu.org> <1482163702.2438.23.camel@tourbillion-technology.com> 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]:58180) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1cJ4zX-0006MY-5q for help-guix@gnu.org; Mon, 19 Dec 2016 15:57:04 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1cJ4zT-0000sA-9q for help-guix@gnu.org; Mon, 19 Dec 2016 15:57:03 -0500 In-Reply-To: <1482163702.2438.23.camel@tourbillion-technology.com> (Paul Garlick's message of "Mon, 19 Dec 2016 16:08:22 +0000") List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: help-guix-bounces+gcggh-help-guix=m.gmane.org@gnu.org Sender: "Help-Guix" To: Paul Garlick Cc: help-guix@gnu.org Paul Garlick skribis: >> > I have noticed that the non-termination problem is not confined to >> > the >> > opencascage_oce package. If I try=C2=A0 >> >=20 >> > $ nix-instantiate --strict --eval --xml -A guile >> > /home/paul/.nix-defexpr/channels/nixpkgs >> >=20 >> > that does not terminate either. >> Does it just take long, or does it simply not terminate? > After a few (approximately five) minutes the process is still producing > xml. =C2=A0Looking at the output, after killing the process, there is a > large amount of blank space plus the occasional line of xml, such as: > path=3D"/nix/store/k3jqgxqfwg6q9bgkys63c25lbjprl3gq-nixpkgs- > 17.03pre96825.497e6d2/nixpkgs/lib/types.nix"> OK. > $=C2=A0./pre-inst-env guix import nix /data/paul/sourceCode/nixpkgs > opencascade_oce > ;;; SSAX warning: Skipping PI: xml > trace: lib.zip is deprecated, use lib.zipAttrsWith instead > trace: `mkStrict' is obsolete; use `mkOverride 0' instead. > trace: `types.list' is deprecated; use `types.listOf' instead > ;; converted from > /data/paul/sourceCode/nixpkgs/pkgs/development/libraries/opencascade/oc > e.nix:5 > (package > =C2=A0 (name "opencascade-oce") Good! > This is almost usable, except the version of the file oce.nix has > changed since the dde259d commit. =C2=A0Also, it may be nix-instantiate d= oes > not terminate for many other packages in recent versions of nixpkgs. Indeed, with Nixpkgs 16.09-beta-6955-g0c7afce, =E2=80=98guix import nix=E2= =80=99 crashes with a stack overflow in the XML parser after a few seconds. Likewise: --8<---------------cut here---------------start------------->8--- $ time NIX_REMOTE=3Ddaemon nix-instantiate -A guile /data/src/nixpkgs --str= ict --eval --xml | wc -c trace: lib.zip is deprecated, use lib.zipAttrsWith instead trace: `mkStrict' is obsolete; use `mkOverride 0' instead. error: stack overflow (possible infinite recursion) 1055021551 real 0m5.959s user 0m5.660s sys 0m0.700s --8<---------------cut here---------------end--------------->8--- IOW, it crashes after having spit 1=C2=A0GB of XML. Woow. :-) Could you report it to the Nix folks? Now, I don=E2=80=99t think =E2=80=98guix import nix=E2=80=99 is that import= ant to be honest. I think its value decreases as the number of packages in Guix increases. Also, all it gives is a rough template and not the harder details; so you may find that packaging OpenCascade won=E2=80=99t take much longer with= out =E2=80=98guix import nix=E2=80=99. Thanks, Ludo=E2=80=99.