The language of the ancients. The workhorse of accounting systems. The technology that runs business. The most popular programming language in the 1960s (and possibly 1970s and 1980s, depending on how one measures popularity).
Where has it gone? More specifically, where have the systems written in COBOL gone?
Demand for COBOL programmers has been declining. Not just this year, but over the better part of the past ten years -- and possibly longer than that. Consider the statistics reported by the site indeed.com:
This link renders an image created in real-time, so your view may differ from mine. But my view shows a steady decrease in the requests for COBOL programmers, from 2006 to 2014.
COBOL was the language used for business systems. It powered accounting systems, payroll systems, invoicing systems, inventory, and reporting for businesses ranging from insurance to telephone to banking to manufacturing. Those businesses are still here (although perhaps spread around the world) and I expect that those accounting systems, payroll systems, and invoicing systems are still with us. (At least, I still get paychecks and bills still arrive.)
I have two theories about this decline:
1) The demand for COBOL remains steady in absolute terms; it is declining as a percentage of the overall technology market.
2) The demand for COBOL is truly declining; the use of COBOL is declining.
With the first theory, systems written in COBOL remain in use. They are maintained as usual, and the demand for COBOL programmers remains steady. The decline of COBOL is a decline in market share, as other languages grow. Certainly the use of C++, C#, and Java have increased over the past decades. One can clearly see the market for Visual Basic in the 1990s. These other languages expand the "pie" and COBOL is growing at a rate less than the market growth.
The second theory is, I believe, the correct one: The demand for COBOL is declining. This can be caused by only one thing: a reduction in the number of COBOL systems.
Systems written in COBOL need a certain amount of maintenance. This maintenance is caused by forces exterior to the system owners: changes in market, changes in technology, and changes in legislation. Thus with a decline of COBOL demand, I can reasonably conclude that there is a decline in the number of COBOL-based systems.
But the functions performed by those old COBOL systems remain. We're still getting bills, in case you hadn't noticed. The bills may be electronic notifications and not paper invoices, but some system is generating them.
Given that the functions are being performed, there must be a system to perform them. (I highly doubt that companies have abandoned computers in favor of people.) Those systems are written in something other than COBOL. The question then becomes... which language?
A better question may be: is there a single language for the business community? Have companies gravitated to a new leader in technology? Or is the business community becoming fragmented?
I've seen nothing of a "new business standard", so I assume that the business community is splitting. The new languages for business are probably C# and Java. Small portions of the business world may be using other languages such as C, Objective-C, JavaScript and node.js, Python, or even F# or Haskell.
Getting back to COBOL. If I'm right about these new languages, then we may have seen "peak COBOL", with the glory of COBOL behind it. That gives us in the industry an opportunity.
The IT age is more than half a century old. We've seen technologies come and go. Not just hardware, but also software. The rise and decline of the big languages (COBOL, BASIC, Visual Basic) may tell us about the rise and decline of other popular technologies, some which are still in the "rise" phase (perhaps tablets).
Perhaps COBOL can still teach us a thing or two.
No comments:
Post a Comment