IBM Z Educator Hub - Group home

z/OS - Learning Journey Digital Mentorship

  

As a highly experienced z/OS tech I know that I do not know everything.  I frequently fail to quickly recall details of things I did know at one time.  During those times, I would check with mentor #1, the internet search engine, etc.

Learning resources and methods are now entering a new era that promises to be just as transformative of a shift from physical books to internet search engines.

ChatGPT <http://openai.com> showcased possibilities and opportunities - frequently referred to as a 'Netscape' moment in the industry.

Right behind was Bard <https://bard.google.com/?hl=en>   

I recently became aware of zOS Search from Living Mainframe <https://zossearch.living-mainframe.de/> 

The above resources are aware of IBM Z, z/OS, and the major z/OS components such as Db2, CICS, RACF, JES2, MQSeries, DFSMS, IMS, etc.

A unique feature of zOS search <https://zossearch.living-mainframe.de/> includes returning 'Sources:' IBM professional manuals with page numbers followed by 'Answer:' which is a written summary/explanation.

The zOS search is useful to those learning z/OS and the highly experienced z/OS technicians.  

Example: A personal friend and very experienced professional working at a customer site contacted me for ideas to isolate and resolve an esoteric and complex issue.  He was struggling with a very deep z/OS environment problem that lacked guidance on how to interpret non-standard diagnostic messages.  zOS search <https://zossearch.living-mainframe.de/> provided the needed problem isolation/resolution guidance.

A quote from the Living Mainframe site, "Recognizing the industry-wide skills gap in mainframe technology, Living Mainframe places a strong emphasis on enabling the growth and development of mainframe skills. They are committed to helping organizations overcome this challenge, ensuring a robust future for businesses reliant on mainframe technology."

My personal opinion is that Bard is superior to ChatGPT for IBM Z and z/OS technical content.  Living mainframe strength at the moment is returning the relevant professional sources of information with page numbers.  Living mainframe currently used ChatGPT for short summary answer.  Bard would be a better choice for the short answer portion of what is returned.

Try these out - Please post your experiences, thoughts, etc.  I would sincerely appreciate and enjoy reading your experiences with the above and your thoughts.

IBM Watsonx is maturing rapidly to compete.  IBM Project Wisdom has a great potential to compete in the technology learning arena.

Paul

Comments

Peter Bishop
Wed May 31, 2023 01:47 AM

Hmmm, it doesn't respond to the Enter key, which is most un-mainframe.

Once you click 'Run' it responds, as you say, with sources etc.

I tried a couple of simple queries, which I found somewhat bland.

Here's a sample comparison with ChatGPT, using the prompt "what is the importance of IEFBR14 to system programmers?"

Living Mainframe answer:
===========begin answer ==========

Sources:

Answer:

IEFBR14 is a program that is used to end a program in a standard way by branching to the address in general register 14. It is not included in z/OS DFSMSdfp Utilities, but it is a useful program that is often used in batch jobs. It is not clear from the given context what specific importance IEFBR14 has to system programmers.

===========end  answer ==========


ChatGPT:

===========begin answer ==========

IEFBR14 is a commonly used utility program in the z/OS operating system. Although it doesn't perform any significant processing, it holds importance for system programmers due to the following reasons:
1.    Dataset Creation: IEFBR14 is often used by system programmers to create empty datasets. It serves as a placeholder for creating empty sequential or partitioned datasets. By specifying the dataset attributes in a job control language (JCL) statement, IEFBR14 can create datasets with the required specifications.
2.    Job Step Filler: In certain cases, system programmers may need to include a job step in a JCL stream simply for organizational purposes or as a placeholder for future development. IEFBR14 can be used as a dummy job step in such scenarios. Since it doesn't perform any processing, it serves as a convenient way to fill the job stream while keeping the JCL syntax intact.
3.    Job Dependency: In complex job flows or dependencies, system programmers may include an IEFBR14 step to act as a predecessor or successor to other job steps. This ensures proper sequencing and coordination of jobs within a job stream.
4.    Program Testing: During program testing, system programmers may utilize IEFBR14 as a temporary solution to bypass certain steps or programs. By replacing a critical program's execution with IEFBR14, system programmers can isolate and debug specific sections of the job without affecting the overall workflow.
5.    Security and Resource Allocation: IEFBR14 can also be used for security and resource allocation purposes. For example, system programmers can employ IEFBR14 to set appropriate security attributes on datasets, restrict access, or allocate resources such as space or buffers.
6.    Documentation and Comments: Including IEFBR14 in JCL streams can also serve as a way to document the purpose or intention of a particular job step. System programmers can add comments or job notes to describe the reason for including IEFBR14, providing clarity for future maintenance or troubleshooting.
While IEFBR14 itself doesn't perform any significant functionality, its versatility and simplicity make it a valuable tool for system programmers when managing and organizing job control language streams, creating empty datasets, maintaining job dependencies, or performing temporary workarounds during testing or development phases.

===========end  answer ==========

Which is better?

So, a good start, but still some work required.

cheers,
Peter