From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mark H Weaver Subject: Re: Our git just broke Date: Sun, 14 Aug 2016 21:08:10 -0400 Message-ID: <87zioealfp.fsf@netris.org> References: <20160813172853.GA11381@thebird.nl> <0d29d381-7127-af6c-0f83-76ac2b0d5b5e@uq.edu.au> <20160814161308.GA1197@jasmine> Mime-Version: 1.0 Content-Type: text/plain Return-path: Received: from eggs.gnu.org ([2001:4830:134:3::10]:58477) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1bZ6P5-0003QH-GT for guix-devel@gnu.org; Sun, 14 Aug 2016 21:09:24 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1bZ6P0-0005xn-6B for guix-devel@gnu.org; Sun, 14 Aug 2016 21:09:23 -0400 Received: from world.peace.net ([50.252.239.5]:47817) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1bZ6P0-0005xg-1x for guix-devel@gnu.org; Sun, 14 Aug 2016 21:09:18 -0400 In-Reply-To: <20160814161308.GA1197@jasmine> (Leo Famulari's message of "Sun, 14 Aug 2016 12:13:08 -0400") 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: Leo Famulari Cc: "guix-devel@gnu.org" Leo Famulari writes: > On Sun, Aug 14, 2016 at 02:28:43PM +1000, Ben Woodcroft wrote: >> On 14/08/16 03:28, Pjotr Prins wrote: >> > Also gnutls does not pass it's tests on my system, nor does >> > subversion. >> Both gnutls and subversion are substituted for me on current master >> (891284), indicating they built without issue OK on hydra. Does it still not >> work for you? > > Regarding GnuTLS, I assume Pjotr is describing the failure of the > "name-constraints" test [0]. > > Unfortunately, GnuTLS used a certificate in one of their tests without > taking care to handle its expiration date, so now the certificate has > expired. It looks like a mistake. Other test certificates in their test > suite are guarded against this, and they have committed a fix. Yes. Unfortunately we cannot fix this with a graft, because that does not prevent the original build from being performed, and if that original build consistently fails after a certain date, I don't see a straightforward way to work around that. While repopulating hydra's store after the failure of its RAID array, I resorted to temporarily setting the system clock back on the mips64el build machine to repeat that build. It might be worth investigating the possibility of setting the system clock to a deterministic value within build containers. Mark