On Mon, 17 Mar 2008 14:21:36 -0600, Howard Brazee wrote:
>On 17 Mar 2008 13:10:18 -0700, (Clark Morris) wrote:
>
>>Although I would question the sanity of anyone who used
>>DISP=(MOD,DELETE) other than with IEFBR14,
>
>Why? I've seen it added to an initial SORT step because that site
>had standards that meant adding a new step meant renaming all of the
>subsequent steps, and documenting the change a bunch of places. I
>didn't question the sanity of the person who did that.
>
Or, simply because I'm lazy and it's one less line of code to
add it to an existing step than code another one.
Can I code DISP=(MOD,DELETE), then DISP=(NEW,CATLG) in the
same step with the same data set name, and assume that order
of processing will first delete and uncatalog the old one,
then catalog the new one? Without SMS? With SMS?
My habit is to code:
//FOO DD DISP=(MOD,CATLG),DSN=my.data.set, ...
//SYSUT2 DD DISP=OLD,DSN=*.FOO,VOL=REF=*.FOO
... to create a data set conditionally and overwrite it if
it pre-existed. I hate JCL.
Someone wrote of "doing the right thing with DISP=(MOD,DELETE)".
What would that be? Defer the recall until the data set is
opened? Other (specify)? But I think it would be great to
be able to code "DISP=(MOD,DELETE),UNIT=(,,DEFER)" explicitly
to achieve the desired effect -- don't mount (i.e. recall)
the data set until/unless it's opened.
-- gil
----------------------------------------------------------------------
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
