by redmail.netbsd.org with SMTP; 4 Nov 1999 13:40:35 -0000
	by gwa2.fe.bosch.de (8.9.1/8.9.1) id OAA18842
	for <tech-security@netbsd.org>; Thu, 4 Nov 1999 14:28:53 +0100 (MET)
	id xma018605; Thu, 4 Nov 99 14:28:24 +0100
          (Netscape Messaging Server 3.0)  with ESMTP id AAA18267
          for <tech-security@netbsd.org>; Thu, 4 Nov 1999 14:40:19 +0100
	by gorgo.bk.bosch.de (8.9.1b+Sun/8.9.1) with ESMTP id OAA10692
	for <tech-security@netbsd.org>; Thu, 4 Nov 1999 14:40:18 +0100 (MET)
Message-ID: <38218CC1.35B9E9C0@de.bosch.com>
Date: Thu, 04 Nov 1999 14:40:17 +0100
From: Guenther Grau <Guenther.Grau@de.bosch.com>
Organization: Bosch Telecom Software Systeme
MIME-Version: 1.0
To: tech-security@netbsd.org
Subject: Re: evil? sshd patch
References: <199911041303.IAA18968@Twig.Rodents.Montreal.QC.CA>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

der Mouse wrote:
> 
> >>> oh, and you'd probably have a lot of trouble interoperating with
> >>> other systems.  :)
> >> Why?  I can't see how this would cause any interoperability
> >> problems.  Or do you know of some wire protocols that care about
> >> what UIDs may be given special powers?
> > Yes, NFS, for example.  It maps userid 0 to anon-userid by default.
> 
> Someone else suggested NFS privately.
> 
> I think you're confusing NFS the wire protocol with certain
> implementations of it.  Some *implementations* map UID 0 to something
> else; this is orthogonal to the *wire protocol*.

I didn't confuse it. I just didn't read your email carefully enough
to get the emphasis on *wire protocol*. I was more focused on
possible interoperability problems in the real world(tm) :-)

Sorry about that,

  Guenther