From Scott Fagen, SHARE Director of Industry Influence 

If there’s one thing that everyone in this country can agree on, it’s that our politics have reached zeniths in polarization and hyperbole. It seems that the little that the government does remains free of this treatment and that includes the mainframe’s position in IT.

Recently, the GAO released a report (GAO-18-298) on the state of affairs of IT at the IRS. The GAO Highlights brief contains this particularly damning statement:

“Mainframes and Servers Services and Support (MSSS) — are facing significant risks due to their reliance on legacy programming languages, outdated hardware, and a shortage of human resources with critical skills. For example, IRS reported that it used assembly language code and Common Business Oriented Language (both developed in the 1950s) for IMF and IDRS, which exposes these investments to a rise in procurement and operating costs, and a decrease in staff available with the proper skill sets. Further, MSSS relies on a significant amount of outdated hardware exposing the investment to rising warranty and maintenance fees, as well as equipment failures.”

I want to let SHARE membership know that we are working on a comprehensive response to this report and will be reaching out for your help in drafting it. Given the public nature of this report and some of the pejorative statements in it, we should expect that IT leaders in our companies will read and react in ways that are likely to be deleterious to our businesses and customers.  

For this month’s Message from SHARE, I wanted to start the conversation by defusing some of the fear, uncertainty and doubt from the statement above.

“…significant risks due to their reliance on legacy programming languages … IRS reported that it used … Common Business Oriented Language (… developed in the 1950s) …”

According to a report by Reuters, there are 220 billion lines of COBOL in use today, 95 percent of ATM transactions, and 80 percent of in-person transactions use COBOL. The COBOL of today is the result of evolutionary change over the last 60 years, including the introduction of features like interoperability with other programming languages and environments (e.g. .NET, Java) and object orientation all while enabling code to continue to run with minimal changes as the language continues to mature.

Could you imagine a GAO report imploring the government to give up on gasoline engine cars because “…of the significant risks due to reliance on legacy engine technologies. Cars rely on engine technology patented in the 1860s …?” While we might want the government to move to more renewables, it’s not because of the age of the underlying technology. The same holds for IT, the desire should be to use the best technology at hand with reasonably managed risk. I can’t see how rewriting the potentially millions of lines of COBOL code into something else, along with all the attendant data and systems management is within anybody’s accepted definition of risk.

If you dig deeper into the report there is evidence that the issue is not with the technology, but with a lack of interest to hire and retain COBOL programmers to keep these applications updated with the latest changes in tax policy. IBM continues to improve the features and performance of the COBOL compiler and these programs, left alone, will continue to run as long as there are machines and data to run them.

“Further, MSSS relies on a significant amount of outdated hardware exposing the investment to rising warranty and maintenance fees, as well as equipment failures.”

In a closer reading of the report, this is really just some literary slight-of-hand to conflate the IBM Z mainframe with equipment from other vendors. As SHARE members are well aware, IBM updates mainframe technology every 18 months to two years with the cost of the machines and maintenance typically reduced (per unit of capacity) with each successive generation.

We will come out with a more comprehensive response in the coming weeks. I encourage you to read the highlights and report and send me your thoughts (sfagen@share.org).