From mboxrd@z Thu Jan 1 00:00:00 1970 From: =?utf-8?Q?Ludovic_Court=C3=A8s?= Subject: Re: SSHv2 doesn't actually have compression levels Date: Thu, 22 Aug 2019 23:48:50 +0200 Message-ID: <87mug0j2gt.fsf@gnu.org> References: <87ef28fflj.fsf@nckx> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Return-path: Received: from eggs.gnu.org ([2001:470:142:3::10]:41937) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1i0uww-0008Fx-0d for guix-devel@gnu.org; Thu, 22 Aug 2019 17:48:58 -0400 In-Reply-To: <87ef28fflj.fsf@nckx> (Tobias Geerinckx-Rice's message of "Tue, 30 Jul 2019 03:58:32 +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: Tobias Geerinckx-Rice Cc: guix-devel Hello, Tobias Geerinckx-Rice skribis: > While reading (guix scripts offload) in a (failed) attempt to get > offloading working, the following comment caught my eye: > > ;; We rely on protocol-level compression from libssh to optimize > large data > ;; transfers. Warn if it's missing. > > which means that we're not currently doing our own compression inside > the tunnel. > > This is significant, because it means that the COMPRESS-LEVEL field of > BUILD-MACHINE is completely bogus[0], and probably always was - unless [...] > [0]: Take a look at OpenSSH's readconf.c, packet.c, and all calls to > start_compression_out(). I don=E2=80=99t get it: the =E2=80=98compression-level=E2=80=99 field defau= lts to 3, and OpenSSH does: static int start_compression_out(struct ssh *ssh, int level) { if (level < 1 || level > 9) return SSH_ERR_INVALID_ARGUMENT; debug("Enabling compression at level %d.", level); [...] } What=E2=80=99s wrong? Ludo=E2=80=99.