Breathe life into older systems

One of the difficulties with the software industry is that in spite of its profile as a supplier of high technology it set its working patterns in the 1980s and has largely stuck with them.

So if a customer requires a software suite, anything from a complete financial solution for an enterprise to a single copy of a word processor, the normal thing to do is to buy a package or an additional license for an extra user and begin using it.

The difficulty increasing numbers of organisations will come across is the nature of the working environment as compared to the requirement for a ‘seat’ in software terms. A seat is clearly a fixed physical spot and to license a piece of software for such a seat and gain some value from doing so, an organisation needs to have someone in that seat for most of the time.

Don't simply bin old IT

In 2007 the pattern of working is changing dramatically. Communications integrator Affiniti conducted research with ICM in November that suggests that by 2012 two thirds of the workforce in the UK at all levels will want the option to work remotely. There will be a need to access critical business applications in many places for a lot of people and the ‘seat’ idea starts to look very dated. Add to this the fact that large organisations can’t simply bin their old technology and start again with something more distributed and flexible and the problem of updating a system but keeping its bare bones emerges as a considerable one.

The solution, as advocated by many software vendors, is to participate in what is known as ‘software oriented architecture’ (SOA). Wikipedia defines this as “a software architecture that defines the use of loosely coupled software services to support the requirements of business processes and software users. Resources on a network in an SOA environment are made available as independent services that can be accessed without knowledge of their underlying platform implementation.”

In other words the IT department takes the existing applications and the newer ones as appropriate and ensures that each individual in an organisation can ‘graze’ as appropriate to whichever task they are performing at a given time. Independent commentators are in favour of it: Steve Cragg, director of analyst organisation Lustratus, welcomes it as a way of deploying older technologies more efficiently than would otherwise be possible.

 “The end result is better value for money from existing investments for longer "

Integration problems

“Because historic investments have often spanned different computer platforms, applications and environments, it has previously been difficult to make these investments work together, but SOA combined with open industry standards addresses this need,” he says. “Also, SOA introduces a means for IT users to realize much greater levels of reuse, with the same sort of cost and quality benefits found in other industries such as manufacturing.” The end result is better value for money from existing investments for longer, he says.

The idea needs careful consideration and implementation, however, if it is to deliver any business benefit. “Implementing SOA tends to increase the complexity of a system's technical architecture by increasing the granularity,” says John Davenport, principle consultant and UK operations manager of Q-Labs, a change management specialist. “For example, in a conventional system one system process performs the business process of handling an order. In an SOA you may need several applets to achieve the same task; input details, validate data, decrement stock, send a dispatch instruction and so on.” That’s complex enough but it isn’t actually achieving anything, he adds. “Just to stand still from a business process perspective this technical slicing and rebuilding of the process must match what the business does. However investing in SOA to replicate an existing process and providing zero tangible gain to the business, even if this is to enable future benefits, is rarely acceptable.”

Dinosaur in lipgloss

Another voice stressing the likely complexity of an SOA solution is Phil Rice, founder and director of the legacy cloning company Erudine. SOA takes old systems and wraps them up in modern dress so that they become usable once more, which is a good thing. “On the positive side, you have a system that is much easier to use, the interface to it has been cleaned up and it can be called a modern service. And if we need to modify the system or add a new piece of functionality, well, we can do this at the wrapper level,” he says. “On the negative side, the root problem of the legacy system has not been addressed. The crumbling system has been coated in a thin layer of paint under the premise that a system that looks nice, is nice.” It may not be, he suggests. “The old system was probably written in an obsolete language, like Cobol, and it’s probably not scaleable. It’s unlikely that it can handle the large throughputs of traffic required for the new web-enabled. You have a dinosaur in lip-gloss.”

Oddly this doesn’t necessarily matter, though. “There are redeeming features to using Service Oriented Architectures, but like all new technology it has to be used with caution,” he says. “Today, Service Oriented Architectures are often hailed as a magic bullet in IT. Yesterday, Object Oriented Architecture. Prior to that it was Business Process Automation”, he says.

It is undoubtedly true that anyone offering any sort of magic bullet for an IT system should be treated with a considerable pinch of proverbial salt. However numerous organisations are dependent on data held in older systems. Fast Internet connections being ubiquitous for on-site staff members and home workers alike, means the idea of wrapping the disparate elements of the corporate system up and presenting them through a service-oriented portal offers a considerable opportunity to address what’s happening in many businesses today.