“We are way, way, way behind Nokia,” said Bill Gates last month. The hitherto all-conquering chairman of Microsoft does not say such things lightly. But the words he uttered next were equally revealing: “We have a long-term view of the wireless market and are making our investments consistent with that.” What he was saying was that ‘Plan A’ – dominate the wireless operating system market – has failed: but that Microsoft has a ‘Plan B’. That alternative strategy is wireless middleware, or the ‘glue’ that helps programs and databases on different computers work together.

Few industries have been hyped quite like the emerging mobile data market. And not without good reason. Wireless applications, mobile computing and wireless Internet access will one day live up to the promise of revolutionising the way that people live and work. One day. Say it quietly, but a ‘new economy’ might even be created.

But this technological nirvana cannot be reached without an investment in a new wave of middleware technologies capable of integrating carriers’ and enterprises’ disparate back-end and front-office systems. The buzzword in this part of the high-tech industry is no longer RoI – return on investment – but ‘RoA’ – return on assets.

This may seem a Herculean task. It certainly sounds like an expensive task – enough to make systems integrators rub their hands together with glee.

But software also promises to solve much of the problem. These new types of middleware come in various forms, including wireless application gateways, service discovery platforms, service delivery platforms, subscriber management systems and device management platforms.

Together, they will form the nucleus around which carriers and enterprises will build their data services. They are also valuable to application developers, who typically have to wrestle with an array of diverse devices. In other words, these particular technologies will make the mobile data revolution possible.

Microsoft’s solution to the problem is the ‘mobile information server’ (MIS) – one of those ‘Trojan horses’ that market researchers like to speculate about.

The idea behind the technology is simple yet compelling: a chunk of middleware that helps with the delivery of Microsoft applications such as Outlook, Excel and Word to mobile devices. But the technology is not as innocuous at it seems. The carrier or enterprise installing MIS locks itself – and, by extension, its customers – into a Microsoft view of the wireless world.

This scenario is not as far-fetched as it seems, and may explain why Gates is keen to play down his company’s prospects. He does not want his potential clients to realise what he is up to. After all, welcoming Microsoft to the

wireless industry’s top table is a strategy that is fraught with danger. Mobile operators’ biggest fear is that they will be mere ‘dumb pipes’, carrying Microsoft-owned content and services to Microsoft end-users. Licensing MIS takes them a big step closer to that.

But Microsoft is not alone in realising the embryonic wireless middleware market’s massive and untapped potential. BEA Systems and IBM were perhaps a little slow in realising the potential of the mobile data industry, but they are now rushing to re-engineer their conventional application servers to meet the burgeoning demand from wireless carriers. Hewlett-Packard and Sun Microsystems, too, are beginning to respond, as are start-ups such as Europe’s Elata and Cognima.

In fact, seasoned industry veterans will by now have noticed something familiar. The names above underline how wireless middleware, like so many other sectors today, boils down to a straight fight between the rival .Net and J2EE [Java 2, enterprise edition] camps. Which one will be swept up in the tornado? Perhaps both.