From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:470:142:3::10]:46322) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1jECQx-0000jo-6M for guix-patches@gnu.org; Tue, 17 Mar 2020 09:39:04 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1jECQw-0006oq-46 for guix-patches@gnu.org; Tue, 17 Mar 2020 09:39:03 -0400 Received: from debbugs.gnu.org ([209.51.188.43]:60553) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_128_CBC_SHA1:16) (Exim 4.71) (envelope-from ) id 1jECQv-0006lo-VN for guix-patches@gnu.org; Tue, 17 Mar 2020 09:39:02 -0400 Received: from Debian-debbugs by debbugs.gnu.org with local (Exim 4.84_2) (envelope-from ) id 1jECQv-0006qh-R5 for guix-patches@gnu.org; Tue, 17 Mar 2020 09:39:01 -0400 Subject: [bug#38546] [PATCH v5] Julia: Update to 1.3.1 Resent-Message-ID: From: =?UTF-8?Q?Nicol=C3=B2?= Balzarotti In-Reply-To: <87zhcfb29s.fsf@devup.no> References: <87imjvwvjx.fsf@guixSD.i-did-not-set--mail-host-address--so-tickle-me> <87fteywndk.fsf@guixSD.i-did-not-set--mail-host-address--so-tickle-me> <87blplwmuk.fsf@guixSD.i-did-not-set--mail-host-address--so-tickle-me> <20200316100641.GK927@E5400> <87fte7772m.fsf@guixSD.i-did-not-set--mail-host-address--so-tickle-me> <20200317081210.GN927@E5400> <87zhcfb29s.fsf@devup.no> Date: Tue, 17 Mar 2020 14:37:53 +0100 Message-ID: <878sjz6r0u.fsf@guixSD.i-did-not-set--mail-host-address--so-tickle-me> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: guix-patches-bounces+kyle=kyleam.com@gnu.org Sender: "Guix-patches" To: Marius Bakke , Efraim Flashner Cc: 38546@debbugs.gnu.org, zimoun Hi, I looked at it. THe patch is literally a one-line change[1] they make to support RFC 3986. According to [2], the patch has been ported to fedora http-parser. Our version is different form the one bundled in libgit2, but deleting the corresponding line (see snippet) libgit2 tests are now passing. However, this http-parser test is failing: *** http_parser_parse_url("http://hostname:/") "proxy empty port" test fail= ed, unexpected rv 0 *** I opened an issue upstream here [3] so that if they fix it we can continue unbundling it without having to patch it ourselves. Hope they decide to upstream it I'll send updates if there are news on the issue, thanks, Nicol=C3=B2. #+name: http-parser.patch #+begin_src diff diff -u a/http_parser.c b/http_parser.c --- a/http_parser.c 2020-03-17 14:05:53.542832960 +0100 +++ b/http_parser.c 2020-03-17 14:05:21.810382406 +0100 @@ -2394,7 +2394,6 @@ case s_http_host_v6_zone_start: case s_http_host_v6_zone: - case s_http_host_port_start: case s_http_userinfo: case s_http_userinfo_start: return 1; default: #+end_src [1] https://github.com/libgit2/libgit2/commit/1bbdec69bef50208f77f0c4cbac7c= 6b56c35973f.patch [2] https://github.com/libgit2/libgit2/issues/5436#issuecomment-593773122 [3] https://github.com/nodejs/http-parser/issues/501 Marius Bakke writes: > Efraim Flashner writes: > >>> In the meanwhile, reding arch bug-tracker [2] I saw that libgit2 depends >>> on http-parser (USE_HTTP_PARSER). Should we unbundle it from git? >> >> That sounds like a separate issue, but a good idea. > > We used to unbundle it up until libgit2 0.99.0, because the release > notes said that their fork is incompatible with the released version of > http-parser. > > We should look into what the compatibility issues are before unbundling > it again. Maybe it's not relevant for Guix.