Why is “Legacy” a Bad Word?

jmckendrick.jpg

A curious case of reverse semantics has existed in the information technology space for some time. That is, the word “legacy”—defined as something handed down from generation to generation—has come to mean outdated, closed, proprietary and all sorts of other nasty things.

Of course, I’m referring to the pejorative term applied to mainframe computers, as well as midrange-class systems, and I guess, at this point, Windows XP, 2000 and NT systems.

Insurance companies, of course, still run many of their operations on these legacy systems. Some companies I have spoken with are running systems dating back to the 1980s.

In the 1990s, I spent a great deal of time in the world of IBM’s AS/400 systems, now called System i. In many ways, these systems are still ahead of their time, and the rest of the world is trying to catch up. More than a decade ago, for example, IBM upgraded the operating system for the platform to 64-bit, which was completely backward compatible and easily put into place with a single installation at the customer site. Last I heard, Intel-based providers were still struggling with the move to 64-bit, and Intel’s Itanium platform took years to get out of the gate. Also more than a decade ago, IBM enabled the machine to be positioned as a high-performance Web server.

In recent years, the mainframe—IBM’s System z —has seen an impressive resurgence. The machine not only runs a mainframe operating system, but also Linux and Unix as well. Applications not only can be run within logical partitions on the main processor itself, which can support thousands of partitions, but also on more lightweight “specialty engines” that don’t use up mainframe cycles, but can take advantage of the mainframe’s speed, reliability and storage capacity.

Just as importantly, the mainframe is fully SOA-enabled, meaning that all applications can be surfaced as services made available to the rest of the enterprise. Needless to say, no one need ever worry about the downtime of those services—there’s no such thing as downtime for a mainframe.

At least 80% of the world’s data still resides on mainframe systems, and this percentage is likely higher for the insurance industry. Drake Coker, writing in Data Center Journal, recently put together a piece on what the world would be like if all the COBOL systems in the world were taken down all at once. We would have no ATMs, no credit card processing, no hotel reservations, no airline reservations, and much, much more. Coker’s point is that while the world keeps trying to tell us COBOL (and by extension, mainframes) are a thing of the past, they actually are the engines that power today’s networked economy.

And, what’s ironic is that the mainframe probably has many times more users these days than it ever did in its heyday back in the 1970s and 80s. At that time, maybe a few hundred people in each organization accessed the system. Now, every consumer and business across the globe, most of the time without realizing it, are bona fide mainframe end-users.

That’s quite a legacy.

Joe McKendrick is an author, consultant, blogger and frequent INN contributor specializing in information technology. He can be reached at joe@mckendrickresearch.com.

The opinions of bloggers on www.insurancenetworking.com do not necessarily reflect those of Insurance Networking News.

For reprint and licensing requests for this article, click here.
Core systems Policy adminstration
MORE FROM DIGITAL INSURANCE