Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!tektronix!reed!percival!nerd
From: nerd@percival.UUCP (Michael Galassi)
Newsgroups: news.software.b,news.admin
Subject: Re: sendbatch broken after patches 9 & 10 (the fix)
Message-ID: <932@percival.UUCP>
Date: Wed, 14-Oct-87 03:48:49 EDT
Article-I.D.: percival.932
Posted: Wed Oct 14 03:48:49 1987
Date-Received: Thu, 15-Oct-87 21:27:58 EDT
References: <465@acornrc.UUCP> <630@gvgpsa.UUCP>
Reply-To: nerd@percival.UUCP (Michael Galassi)
Organization: Percy's UNIX, Portland, OR.
Lines: 37
Keywords: sendbatch, test, 4.2bsd unix
Summary: this is a fix
Xref: mnetor news.software.b:860 news.admin:1170

In article <630@gvgpsa.UUCP> davew@gvgpsa.UUCP (David White) writes:
>In article <465@acornrc.UUCP> bob@acornrc.UUCP (Bob Weissman) writes:
>>The sendbatch script fails after patches 9 & 10 are installed on my
>>4.2bsd system.  The failing line is
>>	while test $? -eq 0 -a \( \( -n "$testit" -s /usr/spool/batch/$rmt \) -o -s /usr/spool/batch/$rmt.work -o  \( -n "$DOIHAVE" -a -s /usr/spool/batch/$rmt.ihave \) \)
>>Which provokes the error message
>>	"test: ) expected"
>>This is strange, since I see balanced parentheses, don't you?
>
>I experienced the same thing under Ultrix 2.0-1 on a uVAX II.  I
>thought it was just me, so I went back to patchlevel 8 since I didn't
>have time to really dig into it.  It looked right to me too.  Any
>suggestions?

I have split this line up for the sake of mailers and all that don't
care for long lines.  The pointed to part is the adition needed to 
make the test work correctly.  By the way, it was patch 10 which
broke things, they were ok before that.  You can also run the old
version of sendbatch with the remainder of news 2.11 at level 10
with no problems.

while test $? -eq 0 -a \( \( -n "$testit"  -a -s /usr/spool/batch/$rmt
                                           ^^^
\) -o -s /usr/spool/batch/$rmt.work -o  \( -n "$DOIHAVE" -a -s
/usr/spool/batch/$rmt.ihave \) \)

The -a (and) was missing, add this and things go smoothly.  I resent
the error message that sh(1) passed back but that is another problem
and is un-related.  I'll have to go to the sources and check on that
much as I resent looking at sh(1)'s source, it is awfull!

-michael
-- 
If my employer knew my opinions he would probably look for another engineer.

	Michael Galassi, Frye Electronics, Tigard, OR
		...!tektronix!reed!percival!nerd