From mboxrd@z Thu Jan 1 00:00:00 1970 From: Danny Milosavljevic Subject: bug#35012: Can't install icecat Date: Wed, 27 Mar 2019 12:14:08 +0100 Message-ID: <20190327121408.114b8ef3@scratchpost.org> References: <87d0md8lya.fsf@posteo.mx> <20190327000941.1aa12a2c@alma-ubu> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha256; boundary="Sig_/GtdNHdKWrTOpUi4xLx0VDtA"; protocol="application/pgp-signature" Return-path: Received: from eggs.gnu.org ([209.51.188.92]:55106) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1h96WP-0005e9-Mk for bug-guix@gnu.org; Wed, 27 Mar 2019 07:15:06 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1h96WO-00014i-Pg for bug-guix@gnu.org; Wed, 27 Mar 2019 07:15:05 -0400 Received: from debbugs.gnu.org ([209.51.188.43]:47359) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_128_CBC_SHA1:16) (Exim 4.71) (envelope-from ) id 1h96WO-00014e-Ky for bug-guix@gnu.org; Wed, 27 Mar 2019 07:15:04 -0400 Received: from Debian-debbugs by debbugs.gnu.org with local (Exim 4.84_2) (envelope-from ) id 1h96WM-0006X9-N2 for bug-guix@gnu.org; Wed, 27 Mar 2019 07:15:04 -0400 Sender: "Debbugs-submit" Resent-Message-ID: In-Reply-To: <20190327000941.1aa12a2c@alma-ubu> List-Id: Bug reports for GNU Guix List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: bug-guix-bounces+gcggb-bug-guix=m.gmane.org@gnu.org Sender: "bug-Guix" To: 35012@debbugs.gnu.org Cc: Alberto EFG --Sig_/GtdNHdKWrTOpUi4xLx0VDtA Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: quoted-printable Just adding a few suggestions for Guix to do: > Especially for big packages like icecat there can always be a gap > between the time when the package definition is updated and when a > substitute is built and ready to be downloaded. Especially since icecat depends on Rust now. Rust has a very long bootstrap chain and if we change something in the base, the chain has to be rebuilt. That takes days on current hardware. Also, currently, we disable tests parallelization in Rust 1.20 because they have an error there. But it would be possible and desireable to enable them for newer versions again and see whether they work. That would entail replacing the "check" phase by one which invokes ./x.py with "-j" (number->string (parallel-job-count)) (instead of the current "-j1" ) And then build like 10 rounds. If it works, the build of each of the corresponding Rust derivations would be faster. --Sig_/GtdNHdKWrTOpUi4xLx0VDtA Content-Type: application/pgp-signature Content-Description: OpenPGP digital signature -----BEGIN PGP SIGNATURE----- iQEzBAEBCAAdFiEEds7GsXJ0tGXALbPZ5xo1VCwwuqUFAlybWwAACgkQ5xo1VCww uqUYDwf/R9J8+NJWMLyKYCpcIhtNQbCP6HXaBHUa76/fiI1wyldFeeLbSE3GBZs6 KzqrXu3c38L8/qNXEjZPwiTPv2HT4zBqUlulmilRofGrojLOwGbauxVR+Kb/rEYI Lb+nF/SyiDM04DaQdK5eJMjKC+xhpZSSlROqcL0+8vMg7bBP3tEHraQaMpKuSwRx 5tdJoV+cKiAzcll7JPdqQ1qNMolGc1Cy6zGANoZhrJ3DLTjrO1pOV8kzEmgqNodj l1gpPbNJJvSBY2EwFoDrH5mR9i2M4Wi2jgGkhqbcdNOJHCaiEof62vKvPWP0HAdt VSQ226XW/kY79+U2RsWCZgqH0nHuVQ== =P2sn -----END PGP SIGNATURE----- --Sig_/GtdNHdKWrTOpUi4xLx0VDtA--