From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from sdaoden.eu (sdaoden.eu [217.144.132.164]) by fantadrom.bsd.lv (OpenSMTPD) with ESMTP id 4f1f0cf7 for ; Thu, 31 Mar 2016 14:24:55 -0500 (EST) Received: by sdaoden.eu (Postfix, from userid 1000) id 7194216045; Thu, 31 Mar 2016 21:24:53 +0200 (CEST) Date: Thu, 31 Mar 2016 21:24:57 +0200 From: Steffen Nurpmeso To: Ingo Schwarze Cc: openssh-unix-dev@mindrot.org, tech@mdocml.bsd.lv Subject: Re: Fwd: Small fixes for ssh.1 and ssh_config.5, OpenSSH_7.2p2 Message-ID: <20160331192457.grb_VWxTg%steffen@sdaoden.eu> References: <20160329192536.QChwcJaIo%steffen@sdaoden.eu> <20160331162003.GF22000@athene.usta.de> <20160331171114.eOBzGfOWc%steffen@sdaoden.eu> <20160331173846.GH22000@athene.usta.de> In-Reply-To: <20160331173846.GH22000@athene.usta.de> Mail-Followup-To: Steffen Nurpmeso , tech@mdocml.bsd.lv, Ingo Schwarze , openssh-unix-dev@mindrot.org User-Agent: s-nail v14.8.8-183-gdd8bc83 OpenPGP: id=95F382CE; url=https://www.sdaoden.eu/openpgp.pub X-Mailinglist: mdocml-tech Reply-To: tech@mdocml.bsd.lv MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Hallo Ingo, list, Ingo Schwarze wrote: |Steffen Nurpmeso wrote on Thu, Mar 31, 2016 at 07:11:14PM +0200: |> I still haven't checked the OpenBSD version of OpenSSH, but i have |> seen your message on their ML (via Gmane) and there you correctly |> state what \& is far, it is used to avoid (mis)interpretation of "!". |> So it should be before that, not thereafter. | |Non sequitur. So now i have, and the bug(s) is (are) still in there. It is a bug, if i type "man ssh_config" i see PermitLocalCommand Allow local command execution via the LocalCommand option= or using the ! Ns command escape sequence in ssh(1). The argum= ent must be ``yes'' or ``no''. The default is ``no''. In OpenBSD language that is total crap. If i type "zcat /usr/share/man/man5/ssh_config.5.gz|mandoc|v" i see ... the very same! |In many languages, you escape characters by placing something before |them. But that isn't true in roff. In some contexts of roff, you can |escape stuff by putting something *after* it. | |>>> and that mandoc falsely does the correct thing, if that is possible. |>> Mandoc correctly does the correct thing. |> The CVS version now does. | |Now you confuse me completely. During the last few years, i changed |nothing in the mandoc code we are talking about. So how can the |version of mandoc matter? The version from September 2015 acted as if "\&! Ns" had been used. I don't have it no more, and am toooo lazy to bite. In the meanwhile i seem to remember a message from A. J. Bentley who reported some bug a while back, did he? I'm drowning in work, i'd surely search otherwise. |>> some people find it more intuitive to put the "\&" afterwards to |>> express that the character is not ending a sentence (i.e. is not a |>> delimiter). | |> A bug. | |No, that practice is just fine. |I don't understand why you keep calling it a bug. Bugbugbugbugbugbugbugbugbugbugbug. (Manually typed.) |> The visible flyspeck in the manuals of the portable version |> springs into the eye and drives you up the wall. Bah. | |I have no idea what you are talking about. Which formatter are you |using? Which output does it generate for which input? Just feeding animals in the zoo. And then home. Tsch=C3=BC=C3=9F. --steffen -- To unsubscribe send an email to tech+unsubscribe@mdocml.bsd.lv