Let the mainframe rule again

I am a retired computer professional who grew up in the early days of mainframes, and lived through the days of mini computers, desktops, and now distributed systems, where everyone has a personal laptop for “end-user” use.

Your recent opinion piece about mainframe utilisation reminded me of the problems experienced during my 40-plus years in and around the industry (Businesses are under-utilising their mainframes).

Two issues stick in my mind. First, that the quality of systems testing exhibited by end users is far worse than that shown by professional IT developers – one illustration being that the number of undetected errors in the average home-grown spreadsheet is about one in 75 formulae included within any given sheet. So attempts by end users to “improve” software can often have undesirable results.

Second, in the days of the mainframe, everything was held in a secured area. Today, much data is held on portable devices that are easily lost. Laptops are also subject to fashion, and thus are expensive to maintain, despite continuing reductions in the “price for power” equation.

The loss of data can be critical to an enterprise. But suppose we say that all critical data should be held in a central secured place, and that it can only be accessed by identified users using identified devices. Result: we don’t lose data. And if the application is also performed in a secure server, users cannot “improve” the software.

Users get a fast service through encrypted Wi-Fi because they send questions to a faster server and get the answers back with minimum disruption. The laptop, in essence, becomes a data terminal device; it does not need to be upgraded every year; and costs are driven down while security improves.

The mainframe becomes king, security and cost/performance is restored, and the laptop returns to its rightful role as a games machine.

Mike Stranks, BCS