On Sun, 16 Jun 2024 21:07:59 -0500, Paul Gilmartin <[email protected]> wrote:

>On Mon, 17 Jun 2024 01:19:27 +0000, Seymour J Metz wrote:
>
>>If you have to unload to 3330, then use BLKSIZE=13030 for SYSLMOD; IEBCOPY 
>>doesn't support COPYMOD reblocking for load/unload.
>> 
>Has everyone noticed what a grievous design/philosophical blunder IBM
>committed by exposing device characteristics to users?  The data should
>appear identical to the user whether they are on a 3330, 339o, 3420-6,
>cards, ... and it should be up to the OS to decide how to put them there.
>OK.  You can't backspace a card punch, but most restrictions beyond
>that are gratuitous.

Note that that is basically what I am doing now.

That triplet - IEWL + IEBCOPY + ftp rdw ends up with effectively
a RECFM=U file that is self-contained and transportable.

On my own OS - z/PDOS - I execute that RECFM=U file directly
rather than needing to go through a load process.

It would be nice if one day that RECFM=U file could be plonked
into any linklib on z/OS and also be executed. And yes, there will
be a slight overhead in doing so - but close to zero chance any
application will be bottlenecked on or be able to detect that
overhead.

I use RECFM=U files for everything. And ignore block boundaries.

And yes, I know that is especially incompatible with existing use
of people using RECFM=U block boundaries to store text (text
meaning written English text or whatever), but by using standard
tools plus just the addition of pdpclib-copyfile, it can all be
converted from/to - which is good enough for me, although
presumably unacceptable to other sites/policies/culture.

BFN. Paul.

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

Reply via email to