The technology world has convinced itself that the next decade will be all about the slow decline of Windows and its somewhat disliked monopoly. What will replace it? Where will the money go? What will Microsoft do if not Windows?

According to IDC’s latest and surprisingly nuanced analysis, while Windows won’t disappear in the next decade, it is on a downward path, eroded by a shift in the model towards mobility, to the cloud, both trends that kill the need for a complex ‘fat client’ like XP, Vista, 7, and for that matter, Apple’s OS X.

In IDC’s view, the next decade will mirror changes in world politics with a plurality of competing powers and world views, read various mobile operating systems (the iPhone, Palm’s webOS, Android, Nokia’s Maemo and perhaps a better Windows Mobile), and once they consolidate from today’s immature and confusing layers (including Microsoft’s Azure), various cloud platforms too.  

PCs will still exist, laptops too, but more and more devices will resemble today’s smartphones and tablets, and there will be a plethora of simpler, low-cost devices, some of which will be based around the only cheap OS of choice, Linux. Power will shift decisively towards transparent mobile platforms that ask as few questions of users as possible to the extent they might not even know what they are using.

I’m not totally convinced that this post-Windows age will be quite as wrinkle-free as this implies, just as the arrival of Internet ubiquity did not necessarily bring forth cheaper services, the death of the middleman, and almost-free communication, as some of its proponents hoped for ten years ago.

Those mobile and cloud ‘platforms’ could just as easily increase cost, duplication and proprietary complexity in other ways - look at the egregiously over-priced iPhone if you doubt that - and  beg the question of why we need ‘platforms’ at all.

There’s nothing new in platforms, they’re where commercial computing started with mainframes more than half a century ago, and while Windows changed the architecture, it hitched people to a platform yoke no less oppressive in that it sought to define and confine choice. You could have the freedom of the Internet as long as you accessed it from a PC running Microsoft software and using, as much as possible, Microsoft applications.

Platforms were and still are about making a bargain. The creator of the platform guarantees applications (the things people actually use) in return for some kind of proprietary tax, be that on the user (Windows), of the device (Intel) or the developer (Apple and its iPhone). Both sides get something out of the deal, or are supposed to.

The arrival a decade ago of environments such as Java, and in a different way the browser itself, was supposed to be about offering freedom from these constraints. Browsers enable access to the apps across platforms, across devices and through services that can exist regardless of underlying software and hardware.

Does the whole platform model still work? Some will say it does, others such as the growing army of social networking junkies, such a debate might sound irrelevant. I bet very few people accessing Facebook today will care whether it was from a PC, a smartphone, or an airport kiosk. They care about the service not the complex web of platforms and technology that enable it. It seems to me that such a world could be held up by much simpler technologies than at present, starting with the browser, Linux, an open-source cloud and paid-for applications and services running on top of all this.

You won’t hear many of today’s big IT companies voting for that of course. They like complexity and tie-ins, and will look to re-create proprietary platform islands in markets such as mobile.

A second issue is how to make computing power accessible to all, not just those able to shell out for a new gadget and software. It is not the battle for dollars, it is the battle for utility, and that goes far beyond the US, Europe and today’s islands of affluence. Computing has become almost as fundamental to economies as energy, but it’s still an expensive resource.

This is why Linux is important in ways that are not just about opposing Microsoft and its products. Linux is a rival not just because it is second option but because its whole model and rasion d’etre is distinct. Linux and the software of the commons doesn’t necessarily lead to innovation per se, but it does help move the cost of it to the parts of the system that people actually use and care about.

Microsoft might worry about its revenue streams a decade hence, but they are not alone in their nervousness. A lot of companies have a lot to lose.