On Sun, 7 Apr 2013 07:06:45 -0500, John McKown wrote:
>
>Personally to use the PARMDD, I would standardize on using VB data sets to
>avoid any problems with sequence numbers and required trailing blanks. This
>still leaves a problem because ISPF Mike's to eliminate trailing blanks.
> 
And ISPF has the misconceived PRESERVE option which preserves not
trailing blanks but line lengths.  If I insert characters it removes
blanks; if I delete characters it adds trailing blanks to compensate.
And there's no way to control or change the number of trailing blanks
in any given line, which should be as simple as using the space bar
or the ERASE EOF key.

>The obvious answer is to  use only UNIX files and  the vi editor to
>maintain them. <grin/>
>
Yup.  Me, too.  Often I generate my JCL with a script.  And as
I've said before, for some instream data sets, I like to use
RECFM=VB,LRECL=254 (an archaic limit imposed by FTP's
JESLRECL).

But there's a problem.  Many IBM utilities still require RECFM=FB,
LRECL=80.  Is there any scheme for JCL which allows both
RECFM=VB,LRECL=254 for my program and in another step
RECFM=FB,LRECL=80 for an IBM utility?  When I tried it with
HLASM, it got I/O errors.  I'd like to be able to code

   //SYSIN  DD  *,RECFM=FB,LRECL=80,

and not get JCL errors.

And I note in the JCL RM that in JES2 catalogued procedures
instream data sets are still limited to 80 bytes (it didn't say
that RECFM=FB is required).  Another instance of halfhearted
implementation.

-- gil

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to