As a computer journalist for the last 25 years, I've received a lot of review copies of software. As something of an obsessive magpie, I've tended to keep most of it, “for reference”. Until yesterday, that is, when I finally threw out all those copies of OS/2, Lotus SmartSuite, and my entire collection of Microsoft software. This included Windows NT 3.5, Windows 2000, Microsoft Office and many, many more. What's makes this little spring-cleaning exercise particularly apt as well as cathartic is that all of us - and not just me - may finally be witnessing the end of the Windows era.

The most obvious manifestation of this has been the utter fiasco that is Vista. This is not so much things like the recent security hole in Vista's memory protection, which may or may not be a big deal – although if Bruce Schneier sees it as a real problem, then I think Microsoft's probably in trouble. No, it's the fact that however Microsoft tries to spin the numbers, Vista is simply not catching on. The company can try all the marketing tricks it likes, but customers are voting with their wallets by opting for the “downgrade” to XP:

An official at PC maker Hewlett-Packard said the majority of the company's business customers are downgrading new computers from Windows Vista to Windows XP -- raising questions about Microsoft's Vista sales figures.

Another tell-tale sign that the days of Windows as we know it are drawing to a close are the increasing number of stories about Midori:

Microsoft is incubating a componentized non-Windows operating system known as Midori, which is being architected from the ground up to tackle challenges that Redmond has determined cannot be met by simply evolving its existing technology.

SD Times has viewed internal Microsoft documents that outline Midori’s proposed design, which is Internet-centric and predicated on the prevalence of connected systems.

Midori is an offshoot of Microsoft Research’s Singularity operating system, the tools and libraries of which are completely managed code. Midori is designed to run directly on native hardware (x86, x64 and ARM), be hosted on the Windows Hyper-V hypervisor, or even be hosted by a Windows process.

That is, we may be looking at a complete dicontinuity in the Windows line – a recognition that the creaking software architecture that has been around for over a decade now is simply too broken to be fixed.

There are a number of straws in the wind indicating that Windows no longer provides the benchmark for desktop computing. The continuing rise in popularity of Apple's machines is a clear indication that more people are looking beyond Microsoft, something that was reflected in Mark Shuttleworth's recent exhortation to the free software community:

"The great task in front of us over the next two years is to lift the experience of the Linux desktop from something that is stable and robust and not so pretty, into something that is art," Shuttleworth said to applause from the audience. "Can we not only emulate, but can we blow right past Apple?"

Quite correctly, Shuttleworth sees Apple as the one to beat these days.

The traditional desktop experience is being further redefined through the move to cloud computing solutions, which are frequently browser-based, and the appearance of GNU/Linux-based ultraportables, which also draw on Net-based solutions for much of their convenience. Both make the underlying operating system moot.

And as if that weren't enough, there's yet another interesting approach that could have a substantial impact on people's perception of free software – and of Windows. The Splashtop technology uses GNU/Linux as a kind of pre-operating system that lets users boot up very quickly to gain Internet access and perform key tasks such as checking email or moving around the Web. Strikingly, some of Splashtop's key advantages are essentially reactions to two of Windows' problems – speed and security:

Be online seconds after you turn on your PC. Why wait for Windows to load when you could be surfing the web right away!

Surf the Web safely, immune from the malware that targets Windows.

What started out as a cool, but rather marginal, idea has now started gain momentum. First, it was announced that Asus motherboards would be offering the feature, and then Asus notebooks. In a sure sign that the approach is entering the mainstream, now Dell has joined the club:

Dell’s Latitude On works by bypassing the Windows operating system so that you get immediate access to things like your calendar, email, Internet, and contacts. It’s a fully integrated technology that will appear later this year on the Dell 4200 and 4300 Latitude series notebooks, and is powered by a Linux OS, sort of as a secondary operating system.

What this means is that users of these systems, even when rigorously Windows-based, will be experiencing the power and speed of GNU/Linux, albeit unwittingly, when they use this mode. Even if they don't know it's GNU/Linux, they will certainly be aware that it is not Windows. Moreover, if the idea proves popular, it's not hard to imagine most PC manufacturers following suit in order to match Dell's move, ensuring that many more people experience the fact that it is possible to use a PC without using Windows.

That is the key breakthrough. At the moment, there is a widespread if tacit assumption that desktop computing *is* Windows. This is manifest most clearly in schools, where teaching computing really means teaching children how to use Windows and Office. Once people realise they have a choice, then Windows' stranglehold on the desktop market will be fatally loosened.

For many years, people in the free software world have dreamed of a day when GNU/Linux would replace Windows on the desktop. Although the market share of GNU/Linux there is finally lifting – figures range from 3% to 11%, according to the sector – it seems unlikely that GNU/Linux will ever take over the desktop from Windows. But that does not mean that Windows will maintain its dominance there, simply that the future is more complex than the monoculture we have seen and suffered for nearly two decades.

I would like to suggest that the free software world should start looking at things from a different perspective – not how many percentage points GNU/Linux gains on the desktop, but how many Microsoft is losing to *all* of the alternatives to Windows. Free software has nothing to fear from a heterogeneous environment – indeed, mixing technologies almost forces open standards upon manufacturers if they want to provide full interoperability.

Moreover, as many have pointed out, the low number of GNU/Linux viruses is only in part down to the latter's superior design. If GNU/Linux became as dominant on the desktop as Windows, there would be a greater incentive to break that security – and a bigger knock-on effect when it was broken. Far better to promote heterogeneity, which is good for the software ecosystem, and thus good for *all* users.

Against this background, then, we should be celebrating what looks increasingly like the end of the Windows era on the desktop, but not in the naïve, and slightly selfish hope that it will usher in one purely based on GNU/Linux. One of the best proofs that free software is superior in all senses to proprietary software is having the maturity to recognise that the world of personal computing today is far more complicated, and far richer, than Bill Gates's original, monopolistic vision of “a (Microsoft-based) computer on every desktop”.

Glyn Moody writes about open source at opendotdotdot.