Cobol users seem to be rather too defensive about their programming language. One of them recently explained to me that a lot of firms couldn't afford to move off Cobol, but tools are available to link from Cobol into the latest technologies, including Java and Microsoft's dot-Net for Web services.

As I thought about it, it occurred to me that people actually don't need to move off Cobol, because it is a good language with a good pedigree. A little more thought and it dawned on me that, actually, far from not needing to move off it, organisations should be very pleased to be running applications written in Cobol.

The language runs on any platform, from a personal computer to Unix and Windows servers to mainframes. It has been used for 40 years and the industry claims that 70 percent of the world's active business applications are still running Cobol code, representing an investment of $5 trillion.

Many of these applications were written long ago, when processors were slow and memory and storage was expensive, using good software engineering principles and a lot of programming discipline.

Those programmers knew how to make efficient use of computer resources. As a result, their code is very fast and very scalable. It has also established a reputation for robustness and security.

And because the code can run on so many hardware platforms, it is easy to migrate the code from one machine to another.

Users of Cobol have enjoyed a rare degree of stability. A number of other languages have emerged during the 40-year history of Cobol. They have promised a lot and then faded away. They have included Assembler, Pascal, ADA, Fortran, Eiffel, SmallTalk and C.

Even now we have C++, Visual Basic, Java and C#. We also have frameworks, such as the Common Object Request Broker Architecture (Corba), Microsoft's dot-Net, and Web services. Cobol has either survived or embraced these technologies, with core application code and its all-important business rules remaining intact.

The big question has long been: how long can Cobol survive? Because it is so portable, it can easily be moved from old platforms onto the latest hardware. There is still a healthy market for tools to manage existing applications and integrate them into Web services and whatever comes along afterwards.

But you have to ask how easy is it to find Cobol skills, in an era when Cobol programming is about as "uncool" as a graduate can get.

My contact says that Cobol skills are in demand, so Cobol programmers find work more easily than their fashion-following colleagues, and Cobol jobs are better paid. Nevertheless, a shortage of Cobol skills seems to be the biggest threat to the long-term future of the language. Though a firm might consider writing a new application in Cobol, the risk of a Cobol skills shortage towards the end of the application's life might persuade them against such a decision.

Firms on newer platforms have got used to throwing higher-spec hardware at their rushed and inefficient code, to increase its performance, especially during the dot-com boom. But in the current economic climate they may find that the funds are not available to continue this strategy.

By comparison those of you still running old, efficient Cobol code are probably laughing. It is serving you well and saving you a fortune as the years roll on. Smile if you are on Cobol.

Have your say: reply to IT Week

More IT Week Comments