Editors note: We shared last week’s blog post, “Creating CICS APIs with HostBridge Mitigates COBOL Skills Shortage for one of Nation’s Largest Banks” through our social media. It elicited this comment from a reader: “Using the technique documented here I do not understand how it solves the problem where a COBOL statement is coded incorrectly. Someone must change the program, compile it and test it.” The reader raised a good point, that I address in this post.
Hello reader… you are absolutely right. If the COBOL program has a bug, there’s no way around it: you gotta squash that bug! And, a COBOL professional is required to do it.
Squashing Bugs versus Enhancing Functionality
However, we see many organizations not just fixing bugs, but struggling to enhance functionality. And, there are a couple of ways to go about this. The organization can certainly make direct code enhancements to their existing COBOL programs. However, in many (most?) organizations judge this to have a high risk profile. Why? Well, in part, because of a “lack of COBOL skills”. But this is where we need to refine our definitions and discussion about the “lack of COBOL skills”.
Understanding the COBOL Skills Gap
Is there really a shortage of people on planet earth that understand, or can learn, the COBOL programming syntax? MAYBE, but I’m not convinced. Is there a shortage of people who understand the (often arcane) COBOL build/test processes that we use under zOS? YES. (But products from Compuware and Broadcom go a long way toward fixing that.) But here’s the real question: Is there a shortage of people who understand the logic of mission critical COBOL programs well enough to make meaningful enhancements? ABSOLUTELY! And that’s the most critical skills gap: the application subject matter experts (SMEs).
It’s the lack of SMEs that raise the risk profile of COBOL enhancement projects. To me, saying that the primary skill of these application SMEs is “COBOL” vastly undersells their value. Their skill is understanding how the a complex set of programs codify business requirements. As it happens, those programs were written in COBOL, perhaps a long time ago. This understanding is the real skills gap causing the premature death of valuable, mission-critical COBOL apps.
- View their existing COBOL apps as components of the overall solution – not the entire solution
- Augment their core mainframe apps with business logic and data sources on different platforms.