Re: Understanding Error Messages

On Mon, 12 May 2008 12:46:38 GMT, "William M. Klein"
<wmklein@[EMAIL PROTECTED]
> wrote:

>Go into HEX mode in ISPF edit of your SOURCE code and look at what is in
column 
>32 of the line with "identification division" on it.  It is a "non-COBOL"

>character (possibly a x"00" or x"0A" or something similar.  Change it to
x'40' 
>(EBCDIC space) and then do the same thing to all characters that aren't
valid 
>COBOL characters as seen in HEX MODE edit of your source code.

All of a sudden the light dawns.  There may be a problem copying out
of the emulator when NULLS ON is in effect.  All fields in the screen
(sequence number in 1 - 6, rest of line 7 - 72, and comment 73 - 80
have all trailing spaces replaced with NULLS (x'00').  If this data is
transferred to a PC file through the emulator via whatever magic weird
things may happen.  Since I have been away from a mainframe since the
end of 2006, I am not quite certain of the details but NULLS ON does
have interesting considerations.
>
>You need to believe your listings.  When they tell you that you have bad
data in 
>a specific column in a specific line, then that is what has happened.
>
>Now, HOW you got the bad data in your source file in the first place is
another 
>question.  Have you had any training on how to use the ISPF editor to
create 
>text files (source code or otherwise)?  If not, you should ask whoever is

>working with you to help you get such training.
>
>P.S.  See:
> 
http://publibz.boulder.ibm.com/cgi-bin/bookmgr_OS390/BOOKS/ISPZEM61/1.3.4.1.3
>
>for how to do an ISPF search (in your source file) for non-visible
characters.
>
>(You probably should review that entire manual if you haven't been
trained on 
>using ISPF for editing)