Thinking laterally I think "Sorting in Rexx" could arguably be better done 
by an interface to DFSORT. But BOTH possibilities are worth asking for.

Martin (who has little to no influence in this)

Martin Packer,
zChampion, Principal Systems Investigator,
Worldwide Cloud & Systems Performance, IBM

+44-7802-245-584

email: [email protected]

Twitter / Facebook IDs: MartinPacker

Blog: 
https://www.ibm.com/developerworks/mydeveloperworks/blogs/MartinPacker

Podcast Series (With Marna Walle): https://developer.ibm.com/tv/mpt/    or 
  
https://itunes.apple.com/gb/podcast/mainframe-performance-topics/id1127943573?mt=2



From:   Rob van der Heij <[email protected]>
To:     [email protected]
Date:   17/11/2017 14:20
Subject:        Re: Pipelines in the z/OS base.
Sent by:        IBM Mainframe Discussion List <[email protected]>



I am happy to see CMS Pipelines discussed here again. I will not try to 
top Hobart's post in length or quality, but some remarks:

- With z/VM 6.4 we renewed the indoor plumbing, replacing the level that 
was 20 years old. We more than doubled the number of built-in programs and 
options (and broke very few pipes in the process).
- One part of the BatchPipes product is BatchPipeWorks - the level of CMS 
Pipelines from 20 years ago. The other part is a set of pipeline stages 
that let two concurrent jobs transfer data with so the consumer can 
already start processing the first part of the output while the producer 
is still producing more output.
- Until 2010 there was a service offering available for CMS/TSO Pipelines 
on z/OS which includes functionality specifically for z/OS and not 
available when running on CMS. Just like some stages only apply to CMS and 
are not available on z/OS. In theory we could have a version on z/OS that 
is very close to what we have with z/VM 6.4 today. I would imagine we'd 
want to merge in the BatchPipes function in some way.
- The NetView Pipes is based on the CMS Pipelines level from the 90's and 
continued to evolve. From what I observe, it has become a rather different 
dialect and it would be a non-trivial effort to merge those paths again 
and have NetView exploit the new function.

The synergy with REXX is in two aspects:
- using PIPE commands in the REXX program to massage data, issue host 
commands and process the output of those commands
- write pipeline subroutines in REXX to create re-usable pieces of code 
breaking up big programming tasks into several small ones
Because pipeline subroutines separate the business logic from the 
environment, it is easy to share code between CMS and TSO.

See 
https://urldefense.proofpoint.com/v2/url?u=https-3A__en.wikipedia.org_wiki_CMS-5FPipelines&d=DwIFaQ&c=jf_iaSHvJObTbx-siA1ZOg&r=BsPGKdq7-Vl8MW2-WOWZjlZ0NwmcFSpQCLphNznBSDQ&m=aJiNuwi_v84--HRy9X4xm_NEIOigcunuqEbElhhFJ7k&s=cZtNr1O5gwSXlzJS9ar3J168WZcs4xtOsVtOVWNQCjc&e=
 
for the key aspects that distinguish CMS Pipelines from UNIX pipes.

When I search the RFE site for "z/OS" it seems that RFE 47699 with 161 
votes is the "hottest" one by far (next is 40 votes). I recall seeing 
another RFE with two dozen votes for sorting in REXX - the plumbers among 
us know you'd get that for free with CMS Pipelines.

Development of new function in CMS Pipelines continues. Without a good 
reason, those enhancements are not developed with z/OS in mind, not tested 
on z/OS, etc. The one-time effort to get CMS Pipelines to z/OS will 
increase.

Sir Rob the Plumber

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




Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 
741598. 
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU

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

Reply via email to