On Mar 20, 2006, at 4:08 PM, David Speake wrote:
I am vague on this but I believe that only CLOSE (if opened system
will close
on ABEND - last decade or two) will write an EOF on DASD, SMS or no
SMS.
Dave,
I am quite sure that it is an SMS option. I can't find the I&T manual
at the moment. We had a few jobs (in my past life) that got "garbage"
when they tried to read a file. It didn't get an i/o error as the
data on that specific track was left over from a previous job. The
"Godzilla" member in parmlib option was quite clear about writing an
EOF marker on all sequential output datasets.
Now they could have deleted the option, I hope not as it was an
invaluable option, IMO.
Maybe Mark Thorem could speak up and tell us if that option is still
available and how it works.
Ed
SMS checks bits in the Format1 DSCB on OPEN INPUT and does not
attempt a
READ. Just gives nice friendly EOF on first READ attempt.
Without SMS - don't even think about it. Worst possible case is if the
first track(s) of the extent contain data left over from a previous
run of the program that supplies your input. Garbage/error reports to
customers instead of nice clean ABEND.
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
