From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ludovic =?UTF-8?Q?Court=C3=A8s?= Subject: bug#33266: guix-copy: Honor the SSH port of a host when defined in ~/.ssh/config Date: Tue, 22 Jan 2019 23:09:07 +0100 Message-ID: <874la0jqf0.fsf@gnu.org> References: <87a7mo5de4.fsf@gmail.com> <87o9b21ehv.fsf@gnu.org> <871s7u53ej.fsf@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Return-path: Received: from eggs.gnu.org ([209.51.188.92]:43507) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1gm4FB-0006ur-SB for bug-guix@gnu.org; Tue, 22 Jan 2019 17:10:06 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1gm4F9-0003xI-DK for bug-guix@gnu.org; Tue, 22 Jan 2019 17:10:04 -0500 Received: from debbugs.gnu.org ([209.51.188.43]:42431) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_128_CBC_SHA1:16) (Exim 4.71) (envelope-from ) id 1gm4F9-0003x9-9N for bug-guix@gnu.org; Tue, 22 Jan 2019 17:10:03 -0500 Received: from Debian-debbugs by debbugs.gnu.org with local (Exim 4.84_2) (envelope-from ) id 1gm4F9-0002iT-1G for bug-guix@gnu.org; Tue, 22 Jan 2019 17:10:03 -0500 Sender: "Debbugs-submit" Resent-Message-ID: In-Reply-To: <871s7u53ej.fsf@gmail.com> (Maxim Cournoyer's message of "Thu, 08 Nov 2018 22:35:32 -0500") 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: Maxim Cournoyer Cc: 33266@debbugs.gnu.org Hi Maxim, Maxim Cournoyer skribis: > ludo@gnu.org (Ludovic Court=C3=A8s) writes: > > [...] > >> This looks good, and indeed it=E2=80=99s an unintended consequence of co= mmit >> cc1dfc202f2fefb6c2eb9467d1fc90a9154550c9. >> >> However, the log of commit cc1dfc had this: >> >> Failing to do that, "%p" would be "0" when using "ProxyCommand" >> in ~/.ssh/config. > > Yeah, I had seen this message, but was doubting if it was still > valid. Helas, it seems so: [...] >> This is arguably a defect in either Guile-SSH or (more likely) libssh, >> and it would be nice to report it. >> >> Would you like to report it upstream? > > It has to be in libssh. I'll try to come up with a repro in C, and > submit the issue to them. With the recent upgrade to libssh 0.8.6, is the issue still present? Ludo=E2=80=99.