As someone who wrote his first program more than 35 years ago, I find it very easy to sound like one of Monty Python’s four Yorkshiremen when I recount how basic computers were in my day. I used to have to load the operating system of my PDP8 using binary dip switches on the front panel. Aye, but we were ’appy!

While the downside of being an old IT war horse is a growing reputation for sending people to sleep, there is an upside: a sense of perspective. A sense of perspective is absolutely essential in IT ­ a business with more fads, trends and ruthless disdain for heretics than any cutting-edge fashion magazine.

Take the latest must-have technology: service-oriented architecture (SOA). I’m not against SOA, but let’s face it, SOA is the distributed computing environment (DCE) in drag. DCE was introduced around 1990 and all of the interoperability aims of SOA are shared with DCE.

The trouble is that DCE didn’t really make it, so why should SOA succeed where DCE failed? Some things, such as the ubiquity of the internet and the arrival of XML, have changed since DCE was on the scene, but most things haven’t really moved on that much. Within the organisation, there has always been interconnection, and a standard such as XML could have been self-enforced so there was no barrier to take-up.

According to the hype, SOA can reduce coding duplication both within the organisation and across the complete business domain, speeding up development, and enabling rapid deployment of sophisticated applications using mash-ups of different services. At the same time, however, SOA can also reduce reliability (the mathematical reliability of a system is the sum of the reliability of its component parts), wreck project timetables by introducing interdependencies, and commit additional expenditure to create services that no one will use.

Like communism, SOA is a great idea, but needs the co-operation of humans to work. Such co-operation is usually hard to obtain, so this is where the risk lies. I can see real benefits from SOA, but it will be hard to achieve them, and the track record of this particular idea is quite poor. In short, I wouldn’t bet my house on the outcome.

I know it’s more exciting (and sells more hardware, software, seminars and journals) to get over-excited about the latest trend. But we shouldn’t parrot what the IT marketers and media preach with the fervour of new religious converts. In my experience, there is room for doubt.

Rather than being fashion victims, we should be like doctors and refuse to get too excited about the latest panacea until clinical trials have proved it effective. And always remember to keep that sense of perspective.

Chris Barling contributes articles for the BCS. He is the chief executive of
e-commerce supplier Actinic. Visit www.bcs.org.uk