From mboxrd@z Thu Jan 1 00:00:00 1970 From: Marius Bakke Subject: Re: [PATCH] gnu: Add kakoune Date: Thu, 26 Jan 2017 23:09:08 +0100 Message-ID: <87vat1324b.fsf@kirby.i-did-not-set--mail-host-address--so-tickle-me> References: <87zijekrns.fsf@zancanaro.id.au> <87y3yykov6.fsf@zancanaro.id.au> <20161230192556.GA2668@jasmine> <87wpehkns6.fsf@zancanaro.id.au> <20170110222611.GA8431@jasmine> <87vat9uqqd.fsf@zancanaro.id.au> <87efzq3yxq.fsf@gnu.org> <87o9yut3b0.fsf@zancanaro.id.au> <878tpx4lrq.fsf@kirby.i-did-not-set--mail-host-address--so-tickle-me> <87y3xx7b0f.fsf@zancanaro.id.au> Mime-Version: 1.0 Content-Type: multipart/signed; boundary="=-=-="; micalg=pgp-sha512; protocol="application/pgp-signature" Return-path: Received: from eggs.gnu.org ([2001:4830:134:3::10]:53692) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1cWsEF-0006Po-UL for guix-devel@gnu.org; Thu, 26 Jan 2017 17:09:20 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1cWsEB-0004Dh-2z for guix-devel@gnu.org; Thu, 26 Jan 2017 17:09:15 -0500 Received: from out3-smtp.messagingengine.com ([66.111.4.27]:52370) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1cWsEA-0004DE-QR for guix-devel@gnu.org; Thu, 26 Jan 2017 17:09:11 -0500 In-Reply-To: <87y3xx7b0f.fsf@zancanaro.id.au> 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: Carlo Zancanaro Cc: guix-devel --=-=-= Content-Type: text/plain Carlo Zancanaro writes: > On Thu, Jan 26 2017, Marius Bakke wrote >> I replaced the patch file with a substitution phase instead ... . > > Is this the preferred way to patch reproducibility issues? I don't > really have a sense of it, but the other packages I have seen have used > either a patch or a snippet to make a build reproducible. > > I don't really have a sense of when I should use a patch, a snippet, or > a phase. Can you give me a sense of when I should prefer one over the > others? When it is a simple s/gzip -f/gzip -f --foo/, I prefer a substitution over a patch since it's easier to maintain/less chance of conflicts. On the other hand, they might silently become ineffective. So perhaps a patch is better. Now, whether this substitution should be run as a source 'snippet' is a different question. The package is reproducible for `guix challenge`, but not for the user verifying this manually with `guix build -S`. I think that is a bug, and the substitution should run as a snippet instead. They are for critical bug fixes, e.g. security, reproducibility or freedom problems. Anything else should go in a phase, AFAIK. Would you like to fix it? Good catch, sorry for making the mistake on your behalf! --=-=-= Content-Type: application/pgp-signature; name="signature.asc" -----BEGIN PGP SIGNATURE----- iQEzBAEBCgAdFiEEu7At3yzq9qgNHeZDoqBt8qM6VPoFAliKc4QACgkQoqBt8qM6 VPovjgf+IynYS/EXfGtLhHWOUioM7MXkkUqJfecdQDGwShW4krQTIDx78lMCWik3 8ZwEQOw8Bidqycz0Gp4NiPObqM+sEWDmX4aXmnYyh6Li1gnGA7GdEjiOlIY6meH8 7QdVNbnkFubm+83gf/H9qQtWzginM/QZidpKVrN0c+kt6o0DY0S3kZg63588uSgI QqxOc56Jpcqy2Rv5p58n1lUwZhIeXWURRuwHl2JfpCcBB4uiWeJpnpkJzVZWsdo3 oVdvqhzvNRffVdWyyPaVFBzDKR5wwsUjcNk6/GoVBqmqKeSZW763wZP5AheIZQnF 3IauKyBL2SMGoiihDBJku969bn2xYg== =7wLQ -----END PGP SIGNATURE----- --=-=-=--