Recently, someone forwarded me a link to an article the Enterprise Tech Journal titled An Open Letter from the Mainframe IT Community: Beware the Bias!
The letter is essentially a response to the U.S. Government Accountability Office’s (GAO) report to the Internal Revenue Service (IRS) in which it suggested that the agency should consider modernizing its mainframe and COBOL based systems.
The GAO reports raises a number of concerns related to the level of effort and expense of mainframe maintenance and operations; risks related to a decreasing population of individuals with appropriate mainframe and COBOL skills; long-term business fit and flexibility. These are the same concerns we hear from clients consistently.
In the spirit of transparency, let’s be clear that Astadia (The Mainframe to Cloud Experts) is a consulting and services firm. We have deep expertise and experience assisting organizations in the modernization of their mainframe-based applications. This almost always means exiting the mainframe itself, but that decision has generally been made before we enter the discussion. Regarding the applications themselves, they could remain in COBOL, be rewritten, retired, or replaced. We only care about getting to the right answer for each client’s unique situation; in fact that is our obligation.
As I read and re-read the “Open Letter” a number of thoughts kept coming to mind:
1. It might be unfair for the group to tightly couple the mainframe and the applications on it. Though that may be part of issue facing organizations, they are two different things and should be treated as such. We often say “the cloud is the new mainframe.” This is not intended to be a cute, throw away line. The cloud has extraordinary compute capacity and resources, with all of the benefits of elasticity.
2. COBOL is not dead, but it isn’t cool. For over thirty years I have told people my responsibility as an IT professional isn’t to be cool, it is to be effective. This is not my generation. Developers can learn COBOL and some are; but a vast majority of digital natives just don’t want to work with COBOL and mainframes. The money can be good and you will employed into the future with less competition – but it just isn’t cool. And by the way, COBOL is not the only language in use causing concern.
3. The “empirical data” argument relative to the better TCO of mainframes (or most things) always concerns me. The old saying that "there are three kinds of lies: lies, damned lies, and statistics" begins echoing in my head. Consulting with clients is a core element of Astadia. We often work with them to understand their specific economics to assess the mainframe relative to other options. Sometimes the mainframe has as good or better economics. However, cost is never the sole decision criteria.
The reality is that the “to modernize or not to modernize” decision is not about infrastructure cost, or human capital, or flexibility, or COBOL, or positioning for needs and capabilities we do not even know about yet. It is about ALL of those issues within an organization’s unique context.
The leadership of each organization, and companies like Astadia, is to have fully formed and fully informed conversations. The only reasonable goal is to arrive at the right answer or answers for each client. And the last thing anyone needs to bring to the discussion is bias and self-serving agendas.