Compuware Total Test brings unit testing to mainframe Cobol development

DevOps environments can use Cobol and mainframes to address their needs

Compuware Total Test brings unit testing to mainframe COBOL development
Credit: IBM

Compuware, a supplier known for its focus on making mainframes a home for today's applications, not a hospice, recently announced an enhancement to its Total Test agile development tool. The goal is to provide DevOps tools to the mainframe world.

Compuware Total Test

Here's what Compuware has to say about Total Test:

  • The world’s largest enterprises run their business on Cobol—but can’t nimbly update that code in response to changing business needs due to quality concerns, antiquated processes and loss of application expertise.
  • Compuware is addressing this critical issue with the first fully automated mainframe unit testing solution that empowers in-house, outsourced, expert and novice developers alike to immediately validate code changes.
  • This Java-like unit testing effectively eliminates the notion of Cobol as “legacy” code that can’t be updated with the same speed and confidence as other code—providing large enterprises with unprecedented agility.
  • The release of Topaz for Total Test underscores Compuware’s leadership of a mainframe renaissance encompassing agile, DevOps, continuous delivery and the generational shift in platform stewardship.

As a result of this innovation, enterprise IT organizations can now ease and accelerate the entire mainframe application lifecycle. Features include:

  • Intuitive visual analysis of even the most complex and poorly documented mainframe programs and data structures (Topaz for Program Analysis and Topaz for Enterprise Data)
  • Real-time quality control and error detection of mainframe coding syntax (Topaz integration with SonarSource)
  • Automated creation and execution of unit tests (Topaz for Total Test)
  • Agile cross-platform source code management and release automation (ISPW and integration with XebiaLabs)

Snapshot analysis

The industry has watched the economics of computing change over time. In the early days, the costs associated with systems, storage and software totally outstripped all other costs. Staff-related costs were way down the list of nearly all cost of ownership (COO) studies conducted by industry research firms. So, enterprises did everything they could do to optimize the use of these very expensive resources, often at the expense of staff related costs.

In the early days, applications were monolithic and complex

This translated to computer languages being designed to support a labor-intensive, slow and deliberate development process. The language compilers were designed to highly optimize the resulting runtime code to reduce the system memory required and the number of machine instructions used to implement to developers' programs to make the best use of the very expensive systems, limited internal memory and very expensive storage systems being used. At this time, compiled languages, such as Cobol, Fortran, PL/1 and a few others were the norm.

Developers spent a great deal of their time working carefully through monolithic applications that were designed to support a complete workload or task. This meant documentation and testing were time-consuming and painful processes.

Today's view is different

Flash forward to today, and we see that systems, memory and storage costs have been overtaken by staff-related, power, cooling and networking costs. Developers are far more likely to use languages that are either interpreted or incremental compilers that are designed to make the staff more productive often at the expense of increased use of processor time and memory. Now languages such as Python, PHP, Ruby or Java are in favor for new development.

Furthermore, development styles have changed. Applications are now constructed of small, easily created and tested, reusable units that implement a specific service lashed together rather than large, complex, monolithic constructs.

Since the developers became one of the largest costs of an IT-based business solution, many tools were created to support agile, rapid development processes for these small units. Often times, however, these tools were designed to support the computer languages that are popular today (Python, PHP, Ruby or Java) and do not support older languages such as Cobol, Fortran or PL/1.

Compuware wants to make 'legacy' languages interesting again

Compuware is one of a small number of suppliers that believes mainframes can address the current requirements to reduce IT costs, increase levels of availability and reliability, and vastly accelerate development to address the ever-more-rapid growth of customer demands for new and enhanced applications. The company believes the proper set of tools, such as its Topaz family, could make Cobol just as relevant to today's business needs as Python, Ruby, PHP or Java.

If your organization still makes heavy use of Cobol-based systems and would like to bring Cobol development out of the 1960s to the 2010s, it would be worth the time to learn more about Compuware and its products.

This article is published as part of the IDG Contributor Network. Want to Join?

Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.
Must read: 10 new UI features coming to Windows 10