From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ricardo Wurmus Subject: Re: [PATCH 0/1] Perl 5.24.0 update Date: Wed, 17 Aug 2016 09:01:51 +0200 Message-ID: <87y43v7uao.fsf@elephly.net> References: <20160816224916.GA22047@jasmine> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit Return-path: Received: from eggs.gnu.org ([2001:4830:134:3::10]:52465) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1bZurX-0007OB-NQ for guix-devel@gnu.org; Wed, 17 Aug 2016 03:02:08 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1bZurP-0006z1-Ip for guix-devel@gnu.org; Wed, 17 Aug 2016 03:02:04 -0400 Received: from sender163-mail.zoho.com ([74.201.84.163]:24554) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1bZurP-0006yx-Ae for guix-devel@gnu.org; Wed, 17 Aug 2016 03:01:59 -0400 In-reply-to: <20160816224916.GA22047@jasmine> 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 Tue, Aug 16, 2016 at 06:43:57PM -0400, Leo Famulari wrote: >> And the perl-no-build-time patch deactivated the code that embedded the >> compilation timestamp that is normally printed by `perl -V`. Perl 5.24.0 >> can take this value from the macro PERL_BUILD_DATE [2], so I removed the >> old patch and added perl-reproducible-build-date.patch. >> >> I don't love this patch... we could instead do something like what >> happens in the Erlang package. There we instead get SOURCE_DATE_EPOCH >> from the environment and do some format string transformations to make a >> pretty date string. > > We could also just keep the perl-no-build-time patch in place. There is > not much point in doing work just to make `perl -V` print a nonsense > date (Unix epoch) IMO. What you sent looks good. It is important to let “perl -V” print a date (even if it is useless) to avoid breaking third-party software that expects a date to be printed. Your patch to just define the date as the epoch seems fine to me. ~~ Ricardo