"Ross Klatte" <klatteross@aol.commmm> wrote in message
news:20040725230705.03961.00001454@mb-m19.aol.com...[color=blue]
> Okay, so he didn't. So, who cares?
>
> The important question is:
> How do I direct my SPUFI output to a disk file?
> I did this before, but now I have forgotten it. (And yet, you say
> Arnold is the dumb one.) If I send SYSPRINT to a disk file, it
> scolds me with "not properly allocated" or what not.
>
> I need a SPUFI attribute that will send output to a disk file.
> If a space allocation is required, I need a DSN description that
> will cover all reasonable (e.g., less than 256 characters and less
> than ten million records) output.
>[/color]
SPUFI automatically writes its output to a sequential dataset, whose name
you specify on the SPUFI panel. This URL gives an illustration of the main
SPUFI panel:
http://publib.boulder.ibm.com/cgi-bi...STHIT#FIRSTHIT
The dataset name of the output file is item 4 ('RESULT') on this page, which
means that the actual dataset name is RESULT with your userid prefixed to
it. For example, if your userid is FRED, the dataset name is FRED.RESULT. If
you want a different name, you can specify it at item 4; if you don't put
the name in quotes, TSO will prefix the dataset name with your userid.
To see the dataset attributes, put a 'Y' in item 5 ('CHANGE DEFAULTS'). You
can change those attributes to any of the values; the page I linked above
shows all of the valid values.
If you prefer to run SPUFI as a batch job rather than using the interactive
panel, simply ask one of your local JCL experts how to code a standard DD
statement for a sequential file and use that for your SPUFI output file.
[color=blue]
> (If you haven't noticed, I have no experience in IBM Mainframe
> programming. You need not be kind, but please be specific.)
>[/color]
Nobody was born knowing mainframe programming or DB2 ;-) We all have to
learn somehow.
Rhino
Arnold Schwarzenegger loses Recall Vote and returns to full-time body building, only to finish 2nd in Mr. Galaxy to Pat Paulsen. 
