Have you noticed that even with all the new technologies available on z/OS, management of the platform has remained the same as it was in the 1980s? Why is that? Is ISPF and batch the pinnacle of user interfaces? Or have we become so comfortable with these interfaces over the decades that we find it hard to change? System programmers are often under the gun and so focused on solving the fire in front of them that they have little time to deal with adopting new technologies. Also, the prevailing support model of the platform has always been, “If it ain’t broke, don’t fix it!” For decades, this has kept people using the same models for managing the system. Why not? It’s always worked.

While this model has worked for people for a long time, it wasn’t always popular. It took quite a bit of IBM investment into ISPF functionality before people adopted it. Our generation of system programmers helped shape ISPF, and now we are facing retirement. The new generation of system programmers do not have decades of experience with this technology. In fact, they have grown up with entirely different technologies and user interfaces, shaping their expectations of how modern systems should interact with users. All of them have experience with both web and Unix command line technologies. We find ourselves at a crossroads between what we have always done and what everyone is doing today. We can try to force new users to use the old technology and blame them for their lack of experience, or we can provide them with interfaces that they recognize, introducing them to a new and more powerful environment. If we allow them to use interfaces they understand, they can start to see the differentiating value that z/OS provides.

You will notice there was no mention of GUIs in the last paragraph. This is because systems management does not rely primarily on GUI tooling. The current management model focuses on tools that enable automation and provides the opportunity for platform resource management to dovetail with resource management across the datacenter. There may be GUI tools used to help people assemble the commands for their automation, while the web technologies can help enhance the efficiency of managing tasks.

Using a different interface does not “dumb down platform management” and it doesn’t mean that people will never understand how things work. In fact, there is no direct correlation between the user interface and understanding of the platform at all. Yes, we are used to using ISPF and batch jobs to manage the platform, but there is nothing magical about that. It is just the way we have always done it. At the heart of our batch jobs is a program, a bunch of inputs, outputs, and perhaps a bit of working area in between.

Imagine managing the system without using ISPF or JCL at all. This is not to say that they will disappear, but we are at a point today where the same things we do in JCL can be done in Python, and the same things we do in ISPF can be done in VSCode or Eclipse. This is not to say that using the old tooling is bad, it just means that using modern tools and techniques is a great way too. Doing it the new way means that new sysprogs can use an interface that is more intuitive for them, making them productive more quickly.

I have been working with a bunch of early tenure mainframers and most of them already have quite a bit of operating system experience. They are comfortable managing Linux environments and understand operating system concepts. They could easily manage cloud and distributed servers, but to manage z/OS they must learn JCL, REXX, and ISPF. Why not remove those barriers, when modern interfaces and tooling are already available. For the moment they may still have to learn a bunch of acronyms (that is a different article), but we need to remove artificial barriers keeping new people from the system and make z/OS easier.

By making management like other platforms, it becomes easier for new people to recognize the differences that matter. Of course, they will still need to learn about the things that make z/OS better. There is no analog to Workload Manager in the cloud. Configuring a sysplex is still unique. There are a bunch of things that make z/OS superior. We need to get our new users to focus on the differences that matter, rather than the ones that don’t. By removing the need to learn about older languages and interfaces at the outset, a new generation of system programmers can manage z/OS resources in a fraction of the time and understand the benefits and magic of the platform sooner (before they get frustrated and quit.)

Making z/OS easier today is not just about providing modern interfaces and tooling to new mainframers. It is about providing a culture that nurtures the latest talent making use of the latest technologies. Those of us that love the mainframe have the golden opportunity to be the agents responsible for inspiring a new generation of mainframe devotees.