Tuesday, September 26, 2006

Are mainframe programmers welcome back? -- Maybe so, and then some

Can I go back to mainframe programming after 4+ years? (The grand plan)

You bet. COBOL and IBM mainframe JCL are verbose, but once you’ve done them for a couple decades, they stay with you forever. Their wordiness is really a kind of self-documentation, which helps long term skill retention.

But as said in earlier entries in this blog, we have to look at the market, and how that comports with personal temperament, in this case, mine. Gradually younger professionals replaced centralized mainframe processing culture with distributed, user or customer driven culture that favored client-server, OOP, and modernism – an languages that actually look cryptic. After the Y2K crunch was over, general demand for mainframe applications developers seem to nosedive for generalists.

At the same time, we saw recruiters troll the country for specialists in orphaned technologies. Although conventional wisdom used to preach diversification as the tocket to job mobility, now it was over-specialization, the willingness to stick with something others perceived as outmoded. The market will live in paradoxes.

“General practitioners” in mainframe technology seemed to become almost unemployable. In-house systems were replaced by specialized packages (like Vantage in life insurance) requiring nitpicky knowledge. Other maintenance operations could be easily offshored.

However since about the beginning of 2006 it seems that there has been an upsurge in demand for mainframe skills. I get many more calls than I used to. I could use better income. I was the super GP, the “family practice” of mainframe computing. Can I go back after 4-1/2 years?

I can be philosophical and claim that the business cycle is a natural, “Darwinian” weeding out, rank-and-yank process. People who drop out are the failures, and have to learn something else, and start over – or, if they are writers and artists, they need to start actually making people pay for their work. Nevertheless, there are a couple of developments that could make mainframe work again.

One is if companies again respect the twenty-plus years of basic big data center experience more than they have recently. I had to develop the maturity of the right way to test, implement and support changes to huge systems that ran huge volumes in every nightly cycle, and deal with the night calls that you did not get paid for (salaried). I had to learn how to make elevations goof-proof. It can be done, but it takes tremendous maturity, judgment, and a “best practices” mentality, as well as absolutely correct use of change control software to guarantee the integrity of the systems. You have to be conscious of the damage you could do. I don’t think that the young hires in a lot of situations have this kind of judgment.

So is the mainframe market for “experienced data center generalists” recovering?

The other situation would occur if the number of “generalist” mainframe positions is actually increasing, and if so many former programmers now in their 50s and 60s have moved on to other things (in my case, trying to get a film made). It used to be a matter of speculation that employers would eventually regret forcing out so many experienced professionals in their 50s with large severance agreements during downsizings. They would need them, and the experience would be gone, especially with older technologies. There seems to be some anecdotal evidence that this is now really going on. Not all offshoring agreements have produced the savings once expected. Market forces, and not just unions, could well reverse the trend to offshore many jobs.

The other possibility for bridging the 4-1/2 year gap is to take certification tests. Brainbench offers certification tests in a wide variety of languages and disciplines. I had taken COBOL, JCL, and ANSI-SQL in 2002, but the certifications expire in three years.

What about advancement?

Well, that is a touchy subject. Let me say right off, that I am not interested in advancement just for the sake of proving that I can “compete” and own a span of control, a domain of people. That does not fit my personality. Companies, when they hire consultants, say they want people to code, test, and implement – they want individual contributors. In many cases, the programmers will be employed by consulting firms that will want to market the contractors publicly as professionals in specialized expertise areas. Normally, that would seem to be the acceptable notion of advancement: to advocate a specific expertise.

Of course, I have been re-inventing myself very publicly in a few controversial areas. Here is not the place to advocate any particular political opinion, other that to express the underlying idea that technology can help the public understand the different modes of the way people think about all kinds of things. Technology can democratize, and it can help reduce the polarization of many political issues that seem to pander to special interests when they are handled by large organizations and lobbying groups, and become partisan.

That’s actually were I would prefer to show leadership. I can think of at least four areas where this gets more specific.

(1) Develop a system to help financial institutions and lenders perform due diligence when granting consumer credit, by checking applications against National Change of Address and performing mandatory consumer notification. The desired end result is to relieve consumers of the worry of having to shred their snail mail.

(2) Develop a cost-effective system to enable webmasters to label all of their content, and standards for all browsers to interpret the content. The end result would be much improved mechanisms to help parents protect children from inappropriate content while allowing adults free expression.

(3) Develop a database application (call it a “do ask do tell” database) to categorize and correlate political and social arguments made on a variety of contentious issues. Call this “knowledge management”, it would build on the open source, Wikipedia paradigm. But the aim is to give high school and college students in social studies tools to develop much more refined critical thinking and debate skills.

All of these projects would require the coordination of the work of a lot of vendors and software companies, with care user and beta testing. For any one project, one company would have to be “in charge.” The projects would require detailed project management. Sorry, I never was a formal project manager. People get certified in project management, too.

Another idea (4), is to develop “best practices” for employer and school-based blogging policies. This has suddenly become a big controversy.

There is a reason that I would want to take leadership ownership on something like one of these ideas, because I have done a lot of research and writing (on my own sites) about all of these since “retirement” at the end of 2001. It all grew out of my 1997 “Do Ask Do Tell” book which grew out of a most subtle political issue. You can read about that at my sites.

Oh, yes, I do want to get the movie made.

No comments: