Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!iuvax!purdue!bu-cs!att!cbnews!mveao
From: mveao@cbnews.ATT.COM (eric.a.olson)
Newsgroups: comp.unix.questions
Subject: Re: Is this a bug in sh?
Message-ID: <7931@cbnews.ATT.COM>
Date: 5 Jul 89 01:50:14 GMT
References: <PAUL.89Jul4155856@marvin.moncam.co.uk>
Reply-To: mveao@cbnews.ATT.COM (eric.a.olson)
Distribution: all
Organization: AT&T Bell Laboratories
Lines: 18

In article <PAUL.89Jul4155856@marvin.moncam.co.uk> paul@moncam.co.uk (Paul Hudson) writes:
>
>Consider the two (executable) shell scripts..
>script1 is		and  	script2 is
> echo $0			echo $0
> name=`basename $0`		name= `basename $0`
>Go on, tell me I haven't RTFM .... ;-) Then tell me why it's this way.

    They both evaluate the basename of their shell -- script1 and script2,
    respectively.  The first assigns this to the variable $name;
    the second, however runs it as yet another command,
    in an environment where the variable $name is set to null.
    This causes recursion and the spawned shells you mentioned.


fodder
fodder
fodder