DD DUMMY an output file?

In the Enterprise COBOL for z/OS 3.3 manual on page 368 there is a statement
regarding OPEN OUTPUT. It says:

Do not specify OUTPUT for files that:
Are defined with a DD dummy card. Unpredictable results can occur.

Huh? I've used DD DUMMY on output files for 25 years and I've never had
unpredictable results. The program always runs and writes to DUMMY, ie. no w
here.
What are they talking about?

On Thu, 30 Sep 2004 22:28:09 -0500, "Ron S" <NoSpam@SpamStopper.org>
wrote:

>In the Enterprise COBOL for z/OS 3.3 manual on page 368 there is a statemen
t
>regarding OPEN OUTPUT. It says:
>
>Do not specify OUTPUT for files that:
>Are defined with a DD dummy card. Unpredictable results can occur.
>
>Huh? I've used DD DUMMY on output files for 25 years and I've never had
>unpredictable results. The program always runs and writes to DUMMY, ie. no 
where.
>What are they talking about?

DUMMY was deemed politically incorrect by the HR department, to guard
the egos of Cobol programmers. The inoffensive replacement word is
NULL.

And your evidence for this MIS-information is what Robert?

The DUMMY parameter is still fully supported in z/OS.  See:
http://publibz.boulder.ibm.com/cgi-.../IEA2B650/12.24

The NULLFILE (not NULL) parameter is described at:
[url]http://publibz.boulder.ibm.com/cgi-bin/bookmgr_OS390/BOOKS/IEA2B650/12.22.2.9[/url
]


--
Bill Klein
wmklein <at> ix.netcom.com
"Robert Wagner" <robert@wagner.net.yourmammaharvests> wrote in message
 news:7vkpl091r3bjgf1vn6ql6ohouuvra78a6q@
4ax.com...
> On Thu, 30 Sep 2004 22:28:09 -0500, "Ron S" <NoSpam@SpamStopper.org>
> wrote:
> 
>
> DUMMY was deemed politically incorrect by the HR department, to guard
> the egos of Cobol programmers. The inoffensive replacement word is
> NULL.
>

> The DUMMY parameter is still fully supported in z/OS.  See:
>  http://publibz.boulder.ibm.com/cgi-.../IEA2B650/12.24
>
> The NULLFILE (not NULL) parameter is described at:
>   http://publibz.boulder.ibm.com/cgi-...2B650/12.22.2.9[/c
olor]

So are you guys implying here that you agree that DD DUMMY produces unpredic
table
results but DSN=NULLFILE does not? What is the deal? Why is the manual sayin
g DUMMY
is unpredictable on output? What the heck is unpredictable about it? Has any
one ever had unpredictable
results? In 25 years I've NEVER seen this fail to do what I expected.

Ron,
I remember the documentation, but don't remember when/why you might have
problems.  I am NOT certain that NULLFILE would be any better.  I have sent 
a
question to one of my "usually" reliable sources for an explanation of the
restriction.

--
Bill Klein
wmklein <at> ix.netcom.com
"Ron S" <NoSpam@SpamStopper.org> wrote in message
news:a4OdnfS0AdnS1cDcRVn_vg@giganews.com... 
>
> So are you guys implying here that you agree that DD DUMMY produces
> unpredictable
> results but DSN=NULLFILE does not? What is the deal? Why is the manual say
ing
> DUMMY
> is unpredictable on output? What the heck is unpredictable about it? Has
> anyone ever had unpredictable
> results? In 25 years I've NEVER seen this fail to do what I expected.
>
>
>

"Ron S" <NoSpam@SpamStopper.org> wrote in message
news:yMWdnbdoFIdQU8HcRVn_vA@giganews.com...
> In the Enterprise COBOL for z/OS 3.3 manual on page 368 there is a
> statement
> regarding OPEN OUTPUT. It says:
>
> Do not specify OUTPUT for files that:
> Are defined with a DD dummy card. Unpredictable results can occur.
>
> Huh? I've used DD DUMMY on output files for 25 years and I've never had
> unpredictable results. The program always runs and writes to DUMMY, ie. no
> where.
> What are they talking about?
The only 'gotcha' I know of is that sometimes you need to code a DCB
parameter in addition to the DUMMY.

> The only 'gotcha' I know of is that sometimes you need to code a DCB param
eter in addition to the DUMMY.

Yeah I know. I've had to do that on output files sometimes, but that hardly 
warrants a "results are
unpredictable" warning in the COBOL manual. If you need the DCB, you code it
 and
the results are perfectly predictable.

Ron S wrote:
> In the Enterprise COBOL for z/OS 3.3 manual on page 368 there is a stateme
nt
> regarding OPEN OUTPUT. It says:
>
> Do not specify OUTPUT for files that:
> Are defined with a DD dummy card. Unpredictable results can occur.
>
> Huh? I've used DD DUMMY on output files for 25 years and I've never had
> unpredictable results. The program always runs and writes to DUMMY, ie. no
 where.
> What are they talking about?
>
>

There should be no problem writing to DD DUMMY, it is the equivalent of
/dev/null in UNIX. (The bit bucket)
So unless things have changed a lot from the old MVS days....

> There should be no problem writing to DD DUMMY, it is the equivalent of /dev/null in UNIX
. (The bit bucket)
> So unless things have changed a lot from the old MVS days....

That's what I thought. But nevertheless there it is in ye ol' COBOL manual.

Ron,
I remember the documentation, but don't remember when/why you might have
problems.  I am NOT certain that NULLFILE would be any better.  I have sent 
a
question to one of my "usually" reliable sources for an explanation of the
restriction.

--
Bill Klein
wmklein <at> ix.netcom.com
"Ron S" <NoSpam@SpamStopper.org> wrote in message
news:a4OdnfS0AdnS1cDcRVn_vg@giganews.com... 
>
> So are you guys implying here that you agree that DD DUMMY produces
> unpredictable
> results but DSN=NULLFILE does not? What is the deal? Why is the manual say
ing
> DUMMY
> is unpredictable on output? What the heck is unpredictable about it? Has
> anyone ever had unpredictable
> results? In 25 years I've NEVER seen this fail to do what I expected.
>
>
>