From owner-freebsd-bugs  Mon Jun  2 02:00:03 1997
Return-Path: <owner-bugs>
Received: (from root@localhost)
          by hub.freebsd.org (8.8.5/8.8.5) id CAA08505
          for bugs-outgoing; Mon, 2 Jun 1997 02:00:03 -0700 (PDT)
Received: (from gnats@localhost)
          by hub.freebsd.org (8.8.5/8.8.5) id CAA08467;
          Mon, 2 Jun 1997 02:00:01 -0700 (PDT)
Date: Mon, 2 Jun 1997 02:00:01 -0700 (PDT)
Message-Id: <199706020900.CAA08467@hub.freebsd.org>
To: freebsd-bugs
Cc: 
From: Harlan Stenn <Harlan.Stenn@pfcs.com>
Subject: Re: conf/3751: Improvements to /etc/rc{,.network,.pccard} 
Reply-To: Harlan Stenn <Harlan.Stenn@pfcs.com>
Sender: owner-bugs@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

The following reply was made to PR conf/3751; it has been noted by GNATS.

From: Harlan Stenn <Harlan.Stenn@pfcs.com>
To: joerg_wunsch@uriah.heep.sax.de (Joerg Wunsch)
Cc: FreeBSD-gnats-submit@FreeBSD.ORG
Subject: Re: conf/3751: Improvements to /etc/rc{,.network,.pccard} 
Date: Mon, 02 Jun 1997 04:56:19 -0300

 >>> On Mon, 2 Jun 1997 08:50:31 +0200, j@uriah.heep.sax.de (J Wunsch) said:
 
 H> -if [ $1x = autobootx ]; then +if [ "$1" = autoboot ]; then
 
 J> That's not entirely confusing syntax.  It's normal habit to be on the
 J> safe side.  Although i agree, the quotes should always be there,
 
 Agreed (obviously)
 
 J> and the X should always be prepended.
 
 I don't think it should ever be there, but if it must be there, it
 should be prepended.
 
 J> Prepending the X saves you from hidden gotchas like in:
 
 J> foo="-ain't this weird?-"
 
 J> if [ "$foo" = "bar" ] ; ...
 
 Did you mean:
 
    if [ $foo = "bar" ] ; ... 
 
 Shell metacharacters aren't expanded with "$foo".
 
 Regardless, if you're talking about the shell metacharacter problem, I'd
 disagree with you about the "hidden gotcha".
 
 Quoting and shell metacharacters are such basic concepts that anybody
 who writes a shell script should be aware of them.  And besides, invoking
 "sh -x script" will quickly show the problem.
 
 As an aside, in "rc" the quoting checking vi.recover is also
 inconsistent...
 
 H