Path: utzoo!mnetor!uunet!steinmetz!dawn!stpeters
From: stpeters@dawn.steinmetz (Dick St.Peters)
Newsgroups: comp.unix.wizards
Subject: Re: Non-standard shell and su.
Message-ID: <8528@steinmetz.steinmetz.UUCP>
Date: 10 Jan 88 20:55:04 GMT
References: <200@icus.UUCP> <264@ho7cad.ATT.COM> <8389@steinmetz.steinmetz.UUCP> <6974@brl-smoke.ARPA> <8400@steinmetz.steinmetz.UUCP> <3333@ulysses.homer.nj.att.com>
Sender: root@steinmetz.steinmetz.UUCP
Reply-To: dawn!stpeters@steinmetz.UUCP (Dick St.Peters)
Organization: General Electric CRD, Schenectady, NY
Lines: 19

In article <3333@ulysses.homer.nj.att.com> ekrell@hector (Eduardo Krell) writes:
>>[my comments on ksh behavior in the presence of symlinks]

>[comments on csh behavior i.t.p.o.s.]

Before we go off on a your-shell/my-shell tangent, remember that this
started when I warned that copying ksh on top of /bin/sh (which
somebody had suggested) could be dangerous.  The point is *not* which
behavior is better (the user should have his/her choice).

If not specifyable, the behavior should at least be predictable.
Having just acquired the task of writing complex installation scripts
for a system whose sheer size necessitates symlinks, I find it a bit
unnerving that /bin/sh behaves differently on different systems.
--
Dick St.Peters                        
GE Corporate R&D, Schenectady, NY
stpeters@ge-crd.arpa              
uunet!steinmetz!stpeters