COBOL's Continuing Value for 3000s
December 5, 2016
The venerable workhorse of COBOL is often maligned during IT strategy meetings. The language usually has to make do with a lot less educational opportunity these days in universities, too. It is verbose and legacy and not ever going to captivate a ping pong table discussion about which platform development platform is best. There are perhaps only 1 million people in the world still trained in using it.
However, COBOL brings a single, enduring asset to the 3000s and other mainframe-caliber servers where it runs. COBOL is a standard, one that's exploited and extended and supported by many vendors, not to mention carried through decades of use. This is no one-man show.
You cannot say that about Powerhouse, or any other fourth generation language. If there's a standard out there for C-Sharp, or the wonders of Visual Basic, it is controlled by a single vendor. Powerhouse users are in a pickle. A single company, Unicom, controls the fate of all users employing the 4GL, and the vendor is jerking its leash on users. When Unicom said it was canceling the license of one customer, James Byrne at Harte & Lyne, then Byrne had a response.
"I am ignoring that," he said. "One cannot cancel a contract without cause."
He went on to point at what sets COBOL apart while he's choosing foundational software. "Not that we use it," Byrne said, "but there is a reason that COBOL is still around. The people who do not understand why are at the root of many of the problems with FOSS." That's open source software he's referencing, something that a vendor cannot cancel, but drifting toward commercial prejudices anyway.
COBOL won't regain its footing at Harte & Lyne, where a 3000 still supports the logistics vendor. But those ping pong developer tools are going out of vogue.
I have pretty much given up on Ruby-on-Rails, having slowly reached the conclusion that the maintainers are more interested in being fashionable than producing a stable and useful product suitable for enterprise deployment.
We will never return to proprietary software, but we are making some significant changes in our approach. We are presently migrating off of Linux onto FreeBSD—and we are moving away from the tech-du-jour crowd onto something more grounded in the commercial world.