Hailing the death cab for Cobol

The Y2k crisis couldn't kill Cobol. But aging Baby Boomers might.

As the world approaches the tenth anniversary of the Y2k crisis, another IT crisis lies on the horizon: the retirement of the Baby Boomer generation. In IT shops, they are the people who care and feed for legacy mainframe applications. And within ten years, most of those people will be gone, says Dale Vecchio, an analyst with Gartner Inc.

"People know that their staffs are eligible for retirement, they know the impact it will have and only just now, within 5-7 years of it being a major impact, are people even talking about it," he says. As with Y2k, IT doesn't want to face up to the problem until until there's an impending catastrophe, he says.

The ultimate loser in all of this will be IBM, Vecchio says. A shortage of people to manage those legacy applications could result in an exodus of enterprise applications from the mainframe.

Less supply, more demand

Not everyone shares that view. Bruce Schneier, chief security technology officer at BT, sees the shortage as a self-correcting problem, not a crisis. It is, he says, a classic Economics 101 supply and demand problem. As the demand for legacy programmers increases, prices will rise and more programmers will enter the field. Sure, a sudden exodus could cause a disruption, but he doesn't see that happening. "It's inconceivable that everybody that knows this skill will retire in the same week without warning," he says. As the number of open positions notches upwards, demand will go up and the market will correct for it, he contends.

But that doesn't mean there won't be pressure -- or that individual businesses in some locations will be able to find new talent they need during this transition period. Stuart McGill, chief technology officer at development tools vendor Micro Focus, agrees that the skills shortage is coming on quickly -- and that everyone won't retire at once. "There isn't a fixed date. This is a constant process of people leaving," he says. But as the pace of retirements quickens, will the market train enough people quickly enough to meet demand?

Cobol programmers: The next generation?

Replacing those Baby Boomers won't be easy. Today, thanks in part to efforts by companies like IBM, there are 75 colleges and universities in the U.S. that teach Cobol. That sounds like a lot -- until you consider that those 75 institutions represent less than 2% of the 4,084+ institutions of higher education in this country.

Furthermore, McGill doubts that today's younger workers will want to learn Cobol. For those up and coming Gen Y-ers, the mainframe is as passe as wearing a wrist watch. "Anyone who is new to mainframe systems is not happy with utilizing the processes and tools around those systems. If you've been trained on Windows using Virtual Studio the last thing you want to do is go back the to mainframe," he says. Of course, all of this is nothing new. Back in 2006, Computerworld correspondent Gary Anthes noted that in a Computerworld survey, 45% of respondents whose organizations use Cobol said their ability to hire Cobol programmers was either "worse" or "much worse" than their ability to hire programmers for modern languages such as Visual Basic, C++ and Java.

But Schneier thinks programmers will follow the money. And with today's modern development environments like Visual Studio and similar tools from vendors like Micro Focus, the transition from modern languages to Cobol will not be too time consuming or difficult for the next generation of programmers to master. "If the market decides that there's a shortage there will be an incentive for people to learn the skills," he says.

Making the problem smaller

There is another way out: Transitioning some or all enterprise applications off of the mainframe. That, however, requires planning that should be happening now, Vecchio says. IT should be conducting risk management of the application portfolio -- but it isn't. Instead IT organizations are waiting until the last minute -- as happened with Y2k, he says.

That can be costly. Poor application portfolio management made tracking down and eradicating the Millennium Bug a huge task because no one knew what programs they had, let alone which were affected. IT got its portfolio in order back in 1998 and 1999, but then let it drop, Vecchio says. Now it's needed again. After a portfolio review, he estimates that 10-12% of a typical company's application portfolio can be eliminated.

The next step is to review what's left and try to get away as much as possible from using that internally developed, legacy code.

Finally, he says, IT needs to bite the bullet and migrate to more modern, lower cost platforms that the next generation is able -- and willing -- to support. "Gen X and Gen Y have greater expectations of what applications will do and they are your future employees," he says. Sooner or later IT must look to the future. Going forward, he says, what these employees ask of IT won't be the same as what people are asking for today. Will the mainframe fit into that picture?

Regardless, he doesn't think businesses will actively manage their application portfolios until the costs of inaction get too high or they see the affects of Baby Boomer retirements. Unfortunately, he says, "We need a crisis to make it happen."

Related Content:

Copyright © 2009 IDG Communications, Inc.

7 inconvenient truths about the hybrid work trend
Shop Tech Products at Amazon