How short are we?
I've heard for years about the mainframe skills shortage and really wondered whether it was actually true, or do companies just not want to pay for skilled MVS people. I recently had occasion to be involved in hiring an MVS systems programmer, and I really think they're right, there is a mainframe skills shortage.
No doubt we are....aging. There's a lot of gray hair at SHARE these days. (Not Fifty shades, but still...) And there seems to be so few people entering the workforce, for various reasons, all of which seems to be being addressed; by IBM, SHARE, academia, etc. But complicating the issue, I believe, is the reluctance of companies to pay top dollar for the skills they want. I took ECON 2 in college...I know the laws of supply and demand. When supply goes down, demand increases, and so does the price.
Don't get me wrong....I've made a very good living over the past 25 years as an MVS systems programmer. But if there's such a skills shortage, why isn't my salary increasing by leaps and bounds? Why don't I see a plethora of six figure job ads out there?
I think we have to temper the 'shortage' talk with the 'cheap' talk.
I do commend the organizations doing something about it: the IBM Academic Initiative, CA's Mainframe Academy, and SHARE itself. I think without these organizations, would truely would be in dire straights.
But I still have to wonder...are we really that short? Or are they just cheap?
How short are we?
Low exposure to the mainframes at high school and college level
I believe big part of the problem is the closed nature of the mainframe industry. In the early days everything mainfreme was essentialy open source and everybody was wellcome to play with it like Linux is these days. Nowadays almost everything mainframe is OCO and there is no easy way to even get a system to play with.
I have been in the mainframe world for 6 years now. And I have not IPLed once. I am an application (assembler) programmer. I can't IPL our development systems. And noone is going to build a system for me to play with. At the same time IBM doe not allow me to run z/OS on an emulator. So I am limited to what I am allowed to do at work. I don't have where to learn just for fun. I have to have a bussiness reason.
If we take the fun out of what we are allowed to do and learn, how can we expect people to be interested?
At the same time there is a lot of stuff you can do and play wih on your own PC. If you dedide you want to build a web server and desing a web page in couple hours you can be on your way and in few days you can have results due to the wast amount of tutorials on the web. You are limited only by your imagination.
If you are a real geek, you can poke around the Linux kernel source code or look at how java or python do garbage collection. It is all open source just two clicks away.
How many clicks away is submitting your first JOB on any z/OS box if you are at high school or college kid? And how many clicks away is doing your first IPL? Six years and still counting for me ...
I believe z/Architecture and z/OS are really sophisticated, exciting, and fun technologies. That's why I am here. But how many young people actually have a chance to figure that out?
SHARE has had an open requirement, SSMVSE04528 Hobby License for z/OS. It received a very high priority in voting. IBM has officially rejected the requirement, but they are working with the MVS Program at SHARE and attempting to understand where the issues are that they are not addressing. Some of the current options are:
RD&T - http://ibm.co/rationalsystemzdevtest (Rational Development and Test Environment for System z)
zPDT System z Personal Development Tool
Unfortunately, both of these options, to my mind, are out of the price range for someone like you or I to use.
The Dallas Systems Center also provides RDP (Remote Development Program). It's a bit lower, but still $550 monthly fee. :(
Suffice it to say, IBM is aware of our (SHARE's) concerns, and are doing their best to address the issues.
Very good blog Mary Ann! I agree that it could be a short/cheap combo that could cause the most pain in the future for the mainframe - for companies and mainframe employees. But in this economy, companies that need mainframers are pinched, just like everyone else. A very similar thing is happening here in Texas, with the shortage of skilled workers in the oil and gas industry. It will be interesting to see what happens over the next 5-10 years. When we had these conversations 5 years ago, it seemed doom was on the way, but I think these days people are working longer, and I know some people have delayed retirement b/c of 2-3x the pay to stay on for a few more years.
Vit, I agree that showcasing z/OS and mainframes to students/newcomers in a more sandbox kind of way would be appealing to many. These days, so many areas of study in college require some sort of programming skills. And I think whichever technology that is easiest to learn, easiest to access, and has the biggest/baddest sandbox wins.