From mboxrd@z Thu Jan 1 00:00:00 1970 From: ludo@gnu.org (Ludovic =?utf-8?Q?Court=C3=A8s?=) Subject: Re: rfc/rfh: i686-w64-mingw32 cross target Date: Thu, 07 Apr 2016 23:12:51 +0200 Message-ID: <87k2k99l7w.fsf@gnu.org> References: <87vb492l7s.fsf@drakenvlieg.flower> <877fgio11v.fsf@elephly.net> <87r3eq2y2s.fsf@drakenvlieg.flower> <8760w2nzl4.fsf@elephly.net> <87d1q8lb7j.fsf@drakenvlieg.flower> 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]:46008) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1aoHEk-0006Ch-AY for guix-devel@gnu.org; Thu, 07 Apr 2016 17:13:11 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1aoHEh-0001Jg-4V for guix-devel@gnu.org; Thu, 07 Apr 2016 17:13:10 -0400 In-Reply-To: <87d1q8lb7j.fsf@drakenvlieg.flower> (Jan Nieuwenhuizen's message of "Sat, 02 Apr 2016 09:30:08 +0200") 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: Jan Nieuwenhuizen Cc: guix-devel@gnu.org Jan Nieuwenhuizen skribis: > Ricardo Wurmus writes: > >> I think these are the only patches I needed. Could you give an example >> of what things need to be patched for mingw? > > The two top commits here are for ncurses and readline > > https://github.com/janneke/guix/tree/wip-hurd+mingw Sorry for the delay, but this is awesome work! I=E2=80=99m impressed that = you got it working in so little time. I=E2=80=99m not offering a full review yet. I think it=E2=80=99s an additi= onal incentive to finally merge the remaining bits from wip-hurd. ;-) My main concern is the complexity of the cross-base stuff. Why is =E2=80=98cross-gcc-core=E2=80=99 needed, for instance? It=E2=80=99s probab= ly inevitable for the complexity to be proportional to the number of target libcs, but I think we should start thinking about how to master this complexity by factoring/abstracting things a bit. A cosmetic comment: I=E2=80=99d put the MinGW packages in a separate module. > (work in progress). Typically: configure flags, adding gnulib modules, > gnulib updates, dealing with non-posix or mingw-specific libc stuff. > > Many more examples here > > https://github.com/janneke/gub/tree/master/patches I think portability patches should be (1) submitted upstream ;-), and (2) applied unconditionally on all architectures (assuming they do not make any difference on the other architectures.) =E2=80=98core-updates=E2=80=99 is open, so now is a good time to add patche= s to the toolchain packages, for instance. :-) Thanks! Ludo=E2=80=99.