Mainframes/COBOL - going to stage a comeback?

[Ganpi Srinivasan]

While engaged in a discussion, a buddy of mine mentioned about some study that in the next 4-5 years there would be a huge demand for Mainframe (COBOL) programmers in the US. Is this a valid conclusion?

I'm wondering if someone in JR had come across that study. If yes, could you please pass it around? I am curious if that situation could, in fact, occur.
Also would it make sense for a Java/J2EE professional to learn COBOL too (apart from a salary/resume boosting perspective).

Please pitch in your thoughts.

Cheers
Ganpi
[ September 20, 2006: Message edited by: Ganpi Srinivasan ]

[Pete Joseph]

If you have the extra time, why not learn cobol, fortran, assembly and so on just for your own knowledge..

But the fact is, there are very specific markets for mainframes and cobol.. I learned Cobol when I was in college, it was interesting to learn, but over my career, it hasn't helped in my career one way or the other.

One of my first career opportunities was working with a Tandem Mainframe system. The problem with mainframe systems are that companies don't always keep up to date with the latest and greatest technologies.

So you might be learning all that you can on a mainframe system called VERSION 4, but if it's 5-6 years old and the latest mainframe system is VERSION 8, then the fact is, unless you have a few friends or buddies who know of those openings, it wouldn't get you another job.

I remember after 4 years working on that tandem system, I was laid off. Half of the people I interviewed with or spoke with had no clue what a Tandem was... The other half laughed at the theory that people still use mainframes.

LIke I said, if you have free time, learn it, but in the long run, don't expect a ton of Cobol jobs or Mainframe jobs to open.. Not when it's cheaper to buy a bunch of servers or to outsource elsewhere..

Mainframes are expensive in the long run and IT has become just another part of business like marketing, sales, accounting and so on.. It's all about the bottom line.

If a cheap server can do the same job as an overpriced mainframe, what is the business going to choose?

And if there are thousands of consultants and workers who know java/j2ee or c#/C++ or .Net while only a select few Cobol programmers out in the world, most businesses aren't going to go against the flow.

Yeah some companies might have a need, but if the latest and greatest is AJAX, then that's what a company wants....not because it works, but because that's what everybody says is needed.

Hey, there might be a pickup in job openings because most of the cobol programmers will be retiring.. If you want to do something different for awhile, go for it, but if you're looking at it in the career sense, just learn it to learn it and don't expect tons of job openings.

[Luke Kolin]

Originally posted by Pete Joseph:
If a cheap server can do the same job as an overpriced mainframe, what is the business going to choose?

The quoted statement above makes a gigantic assumption.

Part of the resilience of mainframes and other "legacy" systems is that they combine extreme age, poor documentation with business criticality. Often times no one really knows the exact tasks these systems do, and therefore cannot easily port the tasks to a "newer" system.

At a recent job the critical transaction management systems of a company were run off of an AS/400, using RPG code that was older than some of the developers I was working with. The workflow required to approve and process a trasnaction had been updated by numerous individuals over three decades, and the increased cost of the machine versus a "cheap server" was nothing compared with the cost of documenting and porting the code, then testing and migrating and retraining many users.

I don't see COBOL and mainframes going away any time soon, especially for high-volume transactions.

Cheers!

Luke

[Pete Joseph]

That's the problem with the legacy systems.. Most have no documentation for them.. I worked on a system where there was no documentation for over 20 years..... It was a nightmare.....

Yeah those systems probably won't go by the wayside, but the fact is, it's not worth just knowing Cobol or any other mainframe these days.. The job market is just not there for that technology.

If you wish to learn it, that's great, but like I said before, I learned java after I worked in mainframes because half of the recruiters and people had no clue what a Tandem system was.

[Jason Cox]

Generally older technologies still have plenty of competition for positions, especially as those positions continue to shrink.

I remember over 10 years ago that people were talking about how we were going to get rid of mainframes and COBOL programmers were going to be in trouble. That transition did not happen overnight like many predicted. Yet at the same time are COBOL jobs growing?

I would not waste my time learning such a language unless I was on a project where I needed to learn enough to convert an app over to something more current or had to know how to interface with it. There are plenty of COBOL programmers on the market already, likely more than there are positions.

Technology is a field that is always moving forward. It doesn't make any sense to go backwards.