Over the decades, COBOL’s enduring success has seen it overcome many challenges. Now, as it reaches 65, should we be making birthday or retirement arrangements? Derek Britton plays the role of party planner.
Timeline – September 1959.
Once upon a time, shortly after Bobby Darin regaled us with the exploits of one Mack the Knife, a quieter cohort met under the auspices of the CODASYL committee. The date, September 1959. The task, agree a name for a new computer language the committee were seeking to develop. No-one quite remembers who suggested it, but the proposal of ‘COBOL’ won the vote, beating other suggestions, including ‘BUSY’, ‘INFOSYL’, and ‘COCOSYL’.
Today, sixty-five years later, while the proponents and die-hard users smile knowingly as it passes yet another milestone, the fact that COBOL has reached traditional retirement age while still extremely successful in the IT world is – to others – nothing short of astonishing.
Still the bedrock application language for many vertical industries and federal and state government departments, COBOL boasts a production footprint of hundreds of billions of lines of code. If you are checking your bank balance, getting an insurance quotation, wiring money, shipping a parcel, booking a vacation or indulging in some digital retail therapy, chances are you’re hitting a COBOL system each time.
As outlined in this article, COBOL’s success is built upon its brilliant design attributes, among them readability, portability, adaptability, and business-centricity. There’s nothing that comes close for business apps.
What is up for debate isn’t that it has been meteorically successful as a technology, more how it achieved it. We’ll explore the issues COBOL has had to tackle, and some of the attributes that have kept it running smoothly for so long. We will draw upon the opinions of experts – as featured in the COBOL 65 guest book – for supporting insight.
The first O doesn’t mean anything
COBOL stands for Common Business Oriented Language and is one of technology’s most enduring technology acronyms. COBOL is a structured high-level language that uses English-like keywords rather than shorter symbols to enable people to learn the syntax quickly. So, the line of code that says ADD A TO B GIVING C does exactly what you think it will (add two numbers and store the result in a third). Relative to many other languages, it is quite hard to misunderstand COBOL, even if you are not a computer scientist.
COBOL’s ease of learning accelerated its widespread adoption. From its first release, in 1960, COBOL became the world’s most widely used computer language inside a decade. It has continued to evolve ever since.
Reasons to be Fearful
We could stop the article there, pretend everything was just fine, and throw the birthday party instead. But COBOL has faced a variety of issues over the years, and existential challenges remain.
Not COOL – COBOL is no longer in Class
The language is dull because its highly structured nature offers nothing to the computer science undergraduate, as there’s little to teach, little to learn, and few employers at the end of the learning process. As such, of the tens of thousands of universities on the planet, COBOL gets airtime in only a few hundred. It has little academic value, is a rather unsightly departure from most other languages, and offers limited direct career prospects. Its hard to see that status quo changing.
The Myth of Maintainability
COBOL became popular really quickly – vast numbers of worker-bee COBOL programmers were building systems as fast as they could go. You can imagine the attention to standards, efficiency and documentation that might have yielded. Sure enough, the effort of maintaining those systems over the years (the systems remained successful – still do in many cases) was far higher than many had hoped.
Why? Well, while COBOL itself is easy to read, an average COBOL application has much more code in it than anyone can easily memorize. A million lines of code is a relatively small COBOL application. Yet a million lines of code would be like trying to memorize a book that is the size of fourteen copies of Tolstoy’s War and Peace. It is just too much information.
While the idea of easily maintainable code remains syntactically true (COBOL is far easier to decipher than symbolic languages such as C or Java), the reality is that a million lines of any computer language is unfathomable without assistance.
The Skills Supply Challenge
Though by no means exclusively, COBOL purportedly suffers from a skills supply issue: the systems in question have outlasted the careers of those who built them. So, there is a second, sometimes third, generation of professionals maintaining systems that were originally developed decades ago. The endless nature of these systems is not matched by sufficient supply of COBOL courses in academia, and many organizations have failed to sustain their ‘bench’ of skilled COBOL professionals. While by no means insurmountable challenges, CIOs will often default to shaping strategy through the lens of available skills. If all your programmers are Java or C#, its hard to argue the case for COBOL from a personnel perspective.
Time’s Up, We’re All Going to the Cloud
COBOL is the long-standing bedfellow of its dear friend CICS and its favoured piece of machinery, the IBM mainframe. Each technology is an elder statesman of computing, peerlessly capable at what they do. Their combined might support the performance, scale and resiliency of many of the global economy’s biggest systems. But when the world and his dog is moving into the cloud, what becomes of those systems? That’s a fundamental challenge facing decades-old COBOL applications today.
The only acronym you need is AI
Artificial intelligence continues to explore the business technology realm as it looks to cement valuable new use-cases to both improve business outcomes and – frankly – to justify its expense. Assisting with the task of ‘legacy system’ maintenance – tools built using a mixture of AI and other technologies offer the modern programmer a new generation of tools to assist them in understanding, reshaping, documenting, changing, testing and even transforming incumbent COBOL systems. What this means for the likelihood of seismic change for said systems remains undefined, but the signs are that it could herald a new wave of change, further undermining COBOL’s credentials.
While other challenges doubtless exist, the list above represent significant and genuine challenges to COBOL’s longer-term viability.
Reasons to be Cheerful
Everything said so far is true, but each point has a counter point. Using the words of others, I want to show the other side of each coin: COBOL’s future promise.
Hip to be Square
Unpopularity is an infamously fickle thing to measure. Where COBOL is taught, the classes are full, and the career path of successful trainees successful. Where it needs to be taught on the job, examples abound of robust COBOL training courses to support the onboarding of new talent – a low-cost means of resolving any skills concerns. Younger generation programmers are happy to pick up COBOL in the numbers that employers need; and adding COBOL to the resume can add a premium to salary. The programs run by the likes of University Professors Cameron Seay and Geoffrey Decker, or by the training organization Interskill Learning, or at the work of Bill Hinshaw’s COBOL Cowboys, suggest that while there remains a demand for talent, channels of supply are extremely active and sought-after.
“COBOL presents an opportunity for educators and students alike because it remains in high demand and is commercially vital while remaining largely overlooked. Students are excited when they learn that COBOL is the powerhouse behind some of the most innovative technologies in the world, and it is a language that is relatively easy to learn with a strong career demand”. Magie Hall, Associate Professor of Information Systems and Quantitative Analysis, University of Nebraska
Maintenance is Strategic
Maintaining core (or ‘legacy’) systems is tough in any language. COBOL typically is where the core ‘must not fail’ systems reside in the IT domain; and as such, risk of failure is a big motivator. Furthermore, inadequate investment in people and tooling and processes has seen associated technical debt rise, making the task of problem solving and systems maintenance comparatively harder.
But it needn’t be. Since even before the Y2K problem, there have been excellent analysis tools for COBOL systems, and the level of innovation for IDEs and other development process tools is on a par with most other languages. The COBOL teams of today have everything at their disposal (budget permitting) to hit their maintenance targets.
“It’s a mistake to focus on the age of your programming language when tackling legacy systems… COBOL can be just as effective a solution for organizations as any other programming language–thousands of large businesses that use it for critical functions demonstrate that. Organizations that think they have a COBOL problem or a mainframe problem, actually have a maintenance problem that they should fix first”. Marianne Belotti, Author of ‘Kill it with Fire.’
Skills is not just a people thing
A gulf exists between organizations who saw the skills gap coming and prepared accordingly, and those who didn’t. But even today, a big gap in talented COBOL staffers is not the end of the world. Fantastic apprentice programs either through academia, training organizations, or in-house or from a partner, or a combination, have and can yield a short-term resolution of any shortfall. And – crucially – the leap forward of technology, again to supplement learning, understanding and documentation of these systems has been tremendous.
“COBOL utilises modern development processes and environments, and happily integrates with other cutting-edge platforms technologies. It’s as modern as everything else – it is truly a language for the ages.” Mike Madden, Owner, Legacy IT
Any Colour You Like
COBOL is a renowned bedfellow of the mainframe. What is less well known is that COBOL has always been – from the outset – a portable language. COBOL predates the IBM System 360 – in fact COBOL was developed across different mainframe systems in the early 1960’s (the first two successfully implemented platforms were RCA and Remington-Rand Univac machines) because there was no market leader back then. COBOL’s portability was part of its design and in later decades, the language rode the tsunami of new operating systems, machine types and chipsets, including, more recently, platforms such as AWS, Azure, and GCP. COBOL works anywhere.
“Modern day COBOL operates efficiently and effectively on virtually any modern operating environment and within literally any Cloud Service Provider.” Craig Marble, Expert Application Modernization Practitioner.
AI is additive
The impact of artificial intelligence on the wider community has been, in principle, to automate menial tasks, and add value to more complex one. With COBOL systems, trundling through millions of lines of data declarations and program syntax can be automated, as can the documentation of what’s found. What it won’t automate is the decision-making over what truly comprises a business rule, or which way to modularize this monolithic program – at least not without expert human intervention and decision-making.
So, while AI is being infused in a variety of technologies that will touch COBOL systems, it is to assist in the efficiency of the delivery – or modernization – of those systems. The same decisions must be made about their evolution, and long-term future. Now, however, they can be made with greater insight. Whether COBOL is part of that long-term plan will be situationally specific. Crucially, genuine insight of the code itself will help make that determination, and a growing array of vendors and service providers offer AI-based tools that can assist with various COBOL modernization activities.
Choosing the Right Party
At 65, some of us are thinking more about our golf swing, or seeing our grandkids more often, or going on that cruise of a lifetime. But COBOL’s cruising along just fine as it is, and while it is no stranger to some of the incumbent difficulties facing those of a certain age, they should not prevent years of future success.
Best placed to offer the most relevant view are, of course, those who use it every day to run their organizations. This example, from one such organization featured in the COBOL 65 Guest Book, sums it up perfectly.
“With COBOL, there is a large footprint of reliance in the Financial, Insurance and Government spaces. It is hard to argue with the value COBOL represents in reliability, performance, maintainability and career opportunities as COBOL has been a part of critical infrastructure for many decades. The syntax is easy to pick up and the routines and procedures therein are the foundations of modern-day automation languages and concepts”. Hector Rivera, VP, Virtualization | Distributed Infrastructure, Fiserv Technology Services.
<< CLICK HERE >> to download a free PDF copy of the full COBOL 65 Guestbook of amazing quotes and anecdotes from end user organisations, consultants, trainers and educators, commentators and experts, and vendors, from around the world.
While there will be reasons to see it change – in places – in the coming years, COBOL appears too pervasive, powerful, and important to be put out to pasture. Its sixty-five-year success is a remarkable achievement. So, let’s throw a 65th birthday party – nobody is retiring just yet, they are too BUSY.
