IBM Mainframe Forum

Hi all,

A quick introduction - I'm new to the mainframe and I'm working in a team of experienced mainframe system programmers and have been learning some of the basics of ISPF/TSO/MVS and I'm looking to expand my knowledge of REXX and utility programming in general. Unfortunatley I often feel that my collegaues have forgotten what it's like to be completely new to the mainframe world and don't understand the best bottom up approach to learning.

I'm trying to be pro-active and get some ideas for people new to the platform and REXX in particular - some useful and interesting things that are worth considering... what are the things that all good sys-progs using REXX should know and should program for themselves....

perhaps some ideas for a simple engaging project that can be used as a good learning tool and is something to build upon - as I know that I will have to train others

for example game ideas/useful utilities that are engaging - and I'd like to be clear in that I'm not asking for code I'd just like some pointers on some good projects that may even have been attempted before that would be good to engage others... as the big barrier to learning any language is not having something to program and a great boon to learning is having a slighly off the wall idea to work on (and some of the IBM redbook material is fairly dry and quite frankly uninspiring).

I realise this is slightly off topic but any advice/tips or hints would be greatly appreciated

I recommend looking at the Rosetta Code page for Rexx.

Hello and welcome to the forum,

If you look here in the forum (Search at the top right of the page) there are Many REXX topics and many contain entire working snippets.

You will be able to find many that you can use for your testing. Just make sure that the Only datasets you use are your own test datasets until you are more familiar.

burn2k wrote:... Unfortunatley I often feel that my collegaues have forgotten what it's like to be completely new to the mainframe world and don't understand the best bottom up approach to learning.

I realise this is slightly off topic but any advice/tips or hints would be greatly appreciated

You're right about us old dinosaurs not being especially understanding of new people. Part of the trouble is TSO, CLIST and Rexx didn't exist in 1968 when I started and the methods we learned in 1968 were batch oriented. Many data set manipulation techniques we learned are now done through TSO, or use IDCAMS if done in batch. In 1968, the use of catalogs was rare, and the tools to analyze catalog issues were primitive to non-existent; in any event none of the methods used in 1968 work in 2013. That's part of the problem.

Another part of the problem is the sheer volume of material a system programmer does now was mostly learned in in a chaotic and unstructured manner. It is difficult for people untrained in structuring unstructured knowledge - most, if not all of us - to present this to new people. For example, in the 1960s and 1970s, system programmers did what were called SYSGENs to prepare systems. Most of the time a SYSGEN was done to update the I/O device configuration. Now this is done through IODF and HCD; MVS establishes the I/O configuration by analyzing the control data established by IODF and HCD, so I/O gens are long forgotten history. The product installation and configuration tasks once performed by SYSGEN are now done by tools like SMP/E. So this knowledge is now 6 feet / 2 meters under.

I hope you appreciate the difficulties you will encounter, as well as the difficulties of the old timer you may question have in responding to your need.

Well, good luck.

Hello,

Something Steve posted reminded me - once upon a time, new system programmers were quite often application developers who showed an interest and an appitude to "systems work". They already had a good working relationship with the "Tech Service" people*. Much of the terminology and mechanics were already familiar even if they didn't have the "license" to do these things.

*No matter how bright someone might be, if they did not work well with the systems programmers, they did not get invited to become one.

If you post things you are unsure of, someone should be able to offer suggestions. All of us were new at one time - even if decades ago. The good news is that many of us still have some working memory

Well, in 1968, the old dinosaurs had maybe 3 years of experience. Us newbies weren't that far behind!

Hi all - thanks for your kind words and responses - I'm glad to find a community of people who work on the mainframe environment and are willing to help.

What I'm looking to build and share is a simple engaging guide that introduces some of the main features of both the REXX language and of the mainframe at the same time. (and in the process learn some more myself) -

My first thought was perhaps a hangman game that reads from a seperate file of words - teaching how to allocate files using TSO ALLOC statements and introducing some of the thinking needed to produce a program from scratch... it would also introduce looping as you'd have to ask for a new letter and a variable that counts the number of trys.

The idea would be to initially plan out the hangman program and just think about the game itself - what are the rules etc....
Then think about the rules in terms of functions - which of the built in functions helps to build the program
What are the problems and how can you build a set of instructions to overcome them...

Can you think of any other simple projects like th idea above that are good for introducing this language to others as I think small thought projects are really great for learning a programming language.

for longer code snippets search for my posts in this forum
and in the <sibling> forum
http://ibmmainframes.com/index.php

thanks Enrico - I'll give that a look always great to find a new resource!