Looking forward to 2010. Part 1 – Killing Windows Mobile to save it.

Analyst and futurist Mark Anderson’s annual predictions often leave you with plenty to think about. He’s one of those people with their finger on the pulse of the world – and not just technology, but economics and science. And even if you don’t agree with him, he’s sown plenty of seeds for discussion and debate.

The first two of this year’s predictions left me with one interesting thought: It’s time for Microsoft to admit that Windows Mobile has reached the end of its life and cancel all current and future development - and possibly to go a whole step further and do the same for Windows CE.

Mark’s predictions are this:

1. 2010 will be The year of Platform Wars: netbooks, cell phones, pads, Cloud standards. Clouds will tend to support the consumer world (Picnik, Amazon), enterprises will continue to build out their own data centers, and Netbook sector growth rates continue to post very large numbers.

2. 2010 will be The year of Operating System Wars: Windows 7 flavors, MacOS, Linux flavors, Symbian, Android, Chrome OS, Nokia Maemo 5. The winners, in order in unit sales: W7, MacOS, Android. W7, ironically, by failure of imagination and by its PC-centric platform, actively clears space for others to take over the OS via mobile platforms.

Why’s this? Well, Anderson suggests that 2010 will be the year that two big technology conflicts come to a head: The Platform Wars and the Operating System Wars. They may sound the same, but they’re very different – platforms are a lot bigger than operating systems, and services like Salesforce.com are on the way to becoming platforms, as is the combination of RIM’s BlackBerry OS, its BES and BIS servers and a whole new generation of web services. The OS Wars have been with us for a long time but they’re starting to coalesce around more than just desktop PCs.

That’s where Microsoft has taken its eye off the ball. With, as Anderson suggests, the success of Windows 7, Microsoft has focused on the desktop PC (and most significantly, its largest customer base: the enterprise desktop). That’s left Windows Mobile neglected, with new entrants rapidly taking its market share. Windows Mobile 6.5 is a set of skins on top of the aging Windows Mobile 6.0 (and the best user experiences still come from handset manufacturers rather than Microsoft), and the promised land of Windows Mobile 7 seems as far away as ever. The Danger acquisition seems to have gone nowhere, with most of the key talent building Android for Google.

The latest iterations of the competing OSes are significantly ahead of Microsoft’s (and how come it’s touting only 800 applications in its Windows Mobile Marketplace compared to tens and hundreds of thousands in its competitor’s online stores, when Windows Mobile used to have one of the most vibrant developer communities of any mobile platform?), and its only credible mobile OS is running on a chipset that appears not to be on the approved list for Windows Mobile 7. That’s the Zune HD, on NVIDIA’s Tegra, with Windows Mobile 7 said to be targeted solely at Qualcomm’s Snapdragon.

So why should Steve Ballmer kill Windows Mobile and Windows CE? The answer goes back to the platform wars. Microsoft talks about three screens plus cloud, a single platform that should go from phone to desktop to TV. It’s meant to be a place where developers can produce code and applications that scale from the smallest devices to the largest displays – using one set of development tools.

And that’s why CE has to go. Yes, you can develop for Windows Mobile in Visual Studio, but both the.NET Compact Framework and the Windows Mobile SDK are cut down versions of their Windows counterparts. If you throw in the convergence between Silverlight and WPF as Microsoft’s user interface strategy matures, it’s clear that Windows Mobile has been left so far behind that there’s no catching up. The three screens are fractured: two of them work, one doesn’t. With developers abandoning Windows Mobile for more profitable pastures, a switch away from a broken strategy stands more of a chance of attracting developers back than anything else (especially if it brings back cool hardware).

But does this mean that Microsoft would be abandoning the mobile space?

No. Microsoft most definitely needs a mobile OS to fulfil it's three screens promise. It just doesn't need the legacy of CE.

What Microsoft needs to do is make everything run Windows. The Windows kernel needs to be ported to ARM, and needs to be the foundation of a whole new mobile strategy. Tools can then cross-compile the same code to different platforms – and .NET can start to fulfil its original cross-platform promise. After all, it’s all about the ecosystem, and the ecosystem needs “Developers, developers, developers”.

That’s not to say that any mobile Windows is going to look like the Windows on your desktop. If anything I’d expect any future ARM-based Windows to be built using the same techniques as Windows Embedded, and resemble the approach taken with the concept Courier notetaking device that was unveiled earlier this year. Taking the Windows embedded approach would let device manufacturers customise hardware and software appropriately, and the modular approach of Windows Embedded (along with the MinWin work that’s already underway) would allow a hypothetical ARM Windows to scale from featurephone to high end smartphone.

There’s another good reason to change Microsoft’s mobile OS. The mobile world itself is changing, shifting to a data-centric model. That approach is at the heart of the next generation of mobile networks, LTE. With LTE network tests starting soon, handset and device manufacturers will be looking for an OS that is based around IP networking, much like Apple’s iPhone port of OS X and Google’s Android. Networking is still really an adjunct to Windows Mobile, as it owes more to its PDA ancestry than to the phone world. As voice calls become VOIP applications, a true mobile Windows is a far better road for the future than Windows Mobile.

Windows CE’s death would mean a loss in the OS Wars. But it would be a loss that could give Microsoft a chance at winning at least some of the Platform Wars.

And that’s just thinking about the first two of Mark Anderson’s predictions. In the rest of our 2009 blog posts we'll be looking at the other 8, and how the'll affect your world in 2010.

--Simon