Windows RT: A Few Conjectures, Based on One Assumption

Windows RT has been in controversy since its launch day. Apart from the debate about whether its existence makes sense, there were two major criticisms. First, since Windows RT was made for ARM tablets, and there for, for touch, why did it include a desktop mode, despite not being able to run legacy desktop software? An obvious explanation is that Windows RT included an Office 2013 suite, which was designed to run in desktop mode. However, this has led to a second criticism: the Office 2013 suite in Windows RT was not touch-friendly. Why didn’t Microsoft design a spacial version of Office 2013 optimized for Metro, and therefore, for touch?

After reading a few articles on Intel’s next-gen Core processors and its Atom SoC on the market, I’ve got a new explanation.

The basic story is, in a post-PC era dominated by ARM tablets like the iPad and Android tablets and smartphones based on ARM SoC, the status of Intel is being challenged. Intel’s x86 based CPU is far more powerful than the best that ARM can offer, but they are also much more power-hungry. However, Intel has been quite successful binning power on its x86 processors: Intel’s Atom based SoC, a 5-year-old architecture, has CPU performance that almost always matches(if not bests) Qualcomm’s daul-core Snapdragon S4, while often consuming less power; Intel is also trying to bring its high-end Core architecture to lower power consumption level, with it unveiling a 7W Ivy Bridge processor on this year’s CES. The current trend of Intel and ARM in the mobile SoC field is that while ARM tries to make its architectures more powerful, Intel is trying to bring down the power consumption of its already-powerful x86 architecture. From what we are seeing now, it is likely that Intel will win, introducing tablets the size and the weight of the 3rd Gen iPad in as early as 2014 that are as powerful as today’s ultrabooks.

And my guess here is: it is because Microsoft has foreseen this future that it decided to give Windows RT an identical interface as Windows 8. Let me explain.

The biggest difference between Intel’s x86 CPUs and ARM’s ARM-based CPU is that the former makes it possible to run traditional Windows programs designed for computers, while the same is impossible on the latter. Also, the biggest criticism about the desktop mode in Windows RT is that, since it cannot run legacy PC software and Microsoft has banned developers from designing applications for the desktop environment in Windows RT, its existence is meaningless. Not only is it meaningless, is also confuses customers: because when they see the desktop mode, they would expect to run programs like Photoshop on it, only to find that they could not.

If Intel was able to reduce the power and heat envelop of its x86 processors so that they match their ARM counterparts in power consumption, there would likely be an performance advantage for Intel’s processors. Combine that with the inert advantage of being able to run legacy software, and you have the recipe for Intel-based Windows tablets to triumph over Windows RT tablets.

Now consider what will happen if Intel-based tablets really dominate the Windows tablets market two years later. Since both Windows 8 and Windows RT include the touch-friendly Metro UI, most Windows tablets will run Windows 8, and most will be able to run legacy software, and they will be as thin and light and longevous as Windows RT tablets. At this time, there will be little value for Windows RT to exist, and Windows 8 and its descendants will take care of the tablet market for Microsoft. The desktop mode will be useful in these tablets, and there will be no more confusion.

However, for Windows tablets to succeed, simply replying on compatibility with legacy software certainly won’t work. Microsoft need good touch experience, which means it will need a large library of Metro apps. It is from here that I am going to talk about the point of Windows RT: attract Metro developers.

Early when Microsoft designed the trio of Windows 8, Windows RT and Windows Phone 8, Microsoft has kept the importance of app ecosystem in mind. The three OSes shares the same kernel, and similar runtime environment for Metro apps, which meant developers could easily port their Metro apps between the three platforms with little or no modification to the code. After their announcement, the biggest controversy in Windows 8 was its inclusion of the Metro Start Screen, while the biggest criticism of Windows RT was its desktop mode. Many believed that the Metro experience won’t work for mouse and keyboard, and that desktop wouldn’t work for fingers. Many believed that Microsoft was using the Start Screen in Windows 8 to lure developers to develop Metro apps, since Windows has such a huge user base, at the risk of alienating all Windows users accustomed to the Start Menu and hated the major change.

That is true. However, if we assume Intel will win in at least the war for Windows tablet SoCs, then the Start Screen touch-friendliness will be put to use. Microsoft is not only doing this for Metro; it is taking the first step towards PC-tablet convergence with the Start Screen in Windows 8.

So where do Windows RT fit into this? As we can see now, Windows 8 based tablets have begun to hit the market, like the Acer W700 or Microsoft’s own Surface Pro, but from a tablet perspective they are far from competitive: they require fans, they are bulky, and they suffer from abysmal battery life compared the likes of the iPad. Microsoft may want the convergence and Wintel tablets to happen, but it needs a stop-gap solution, and here comes the Windows RT.

Metro is a great idea. It is a great touch interface, and it offers a fresh experience unmatched by the competition. Windows RT tablets have the potential to gain momentum in the tablet market, which means with Windows RT Microsoft stands chance to get a foothold there. As long as there is a user base, there will be developers making apps, and the system will grow. Waiting for Intel to catch up with ARM may result in a situation where it would be too late for Microsoft to come back, but with Windows RT Microsoft can still manage to stay relevant.

So my guess is, the purpose of Windows RT is to let developers see hope in Metro and Microsoft, and therefore have the incentive to build up the ecosystem together with Microsoft. The Start Screen in Windows 8 isn’t going away, the one that will finally go away is Windows RT itself.

This didn’t answer why Microsoft included a desktop Office 2013 suite and a desktop mode in Windows RT. Again, allow me explain.

I think there are two reasons for these, the first being the ease of transition, and the second being Microsoft’s understanding for what a ‘tablet experience’ is for.

The first was easier to understand. Everyone who has used Windows 8 knows the shock when they are unexpectedly thrown into Metro from the desktop. If Windows 8 on tablets is the future, this isn’t going away. Also, to use Windows 8 on a tablet, you have to realize that a desktop mode exist, and whenever necessary, you can utilize the desktop environment.

If Windows RT is the stop-gap solution, Microsoft must make sure that early adopters of Windows tablets are used to these quirks in future Windows tablets. Hence the desktop environment in Windows RT. In fact, by enforcing the Metro Start Screen on Windows 8 users, Microsoft is also educating traditional form factor users in preparation for the convergence in the future. When switching between Metro and desktop is no longer a pain, well, you are ready to use hybrid devices, like the Surface Pro.

As for the second reason, it has been commonly accepted that tablets are content consumption devices, which means they are used for entertainment and light work, not for serious productivity. If you look at the iOS App Store, you will see that trend. If you want the best example of this philosophy, I’ll point you to Amazon’s Kindle Fire.

In crafting an Office 2013 suite for Windows RT, Microsoft has two choices: simply leave it as it is and port it to ARM, or design a touch-optimized version from ground up.

If Microsoft chose the latter, the first thing it would need to do is to clear up the interface, so that touch targets would get bigger. This means reducing menu entries, which almost certainly means reducing functions. Otherwise, the depth of each menu would grow, resulting in a interface that’s too complexed to be intuitive: consider for example 16 options, now dispersed evenly in 4 menus entries, that means each entry has 4 items; should we reduce the menu entry to 2, then each entry will contain 8 items. Office is far more comprehensive, and therefore, clearing up the interface almost becomes a mission impossible.

So Microsoft would need to strip down Office should it choose the second route. Then, the advantage of the second route suddenly become vague. You want serious work done on your tablet? Surely a stripped down Office won’t satisfy you, and who would be typing an essay using on-screen keyboard anyway? You want casual editing? Guess you can manage that with the on-screen keyboard and a desktop-like Office 2013.

Taking into consideration that tablets are traditionally content consumption devices, I see the reason why Microsoft simply ported Office to ARM. Microsoft also offered a solution for customers coming after productivity in its own tablets: the Touch Cover and the Type Cover. With either, you can get serious things done in serious Office.

Despite touting Surface as productivity devices, from this solution I get the feeling that Microsoft sees tablets in this way: it’s hard for true productivity to work for touch, so productivity will remain in the desktop; as for content consumption, just go to Metro.

This will be more obvious when x86 based Windows tablets become the mainstream: you can code in desktop, use Photoshop in desktop, use Office in desktop, whatever you want; you can play Angry Birds in Metro, read books in Metro, watch movies in Metro, whatever you like. Tablet’s will contain both work and play, and they will be divided in two environments, and in each environment you get the maximum functionality.

So this, I guess, is Microsoft’s vision for the post-PC era:

Desktop will not go away. Neither will Metro. The two will co-exist in every device running Windows with screens larger than 7 inches. Tablets running Windows will be as thin and light and longevous as the iPad, but as powerful as ultrabooks we are seeing today. You can use a tablet for everything, or a laptop, or both. If you are a casual user, you go with the tablet. When you go for consumption, you use Metro apps; when you go for productivity, you plug in a keyboard of some form and fire up the traditional productivity software. You can also use touch to navigate the desktop, should you so inclined.

This is a wonderful conjecture, the true post-PC world, where tablets really become competent laptop replacements. Should this come into reality, Microsoft will once again dominate the market: large as the app library of the iPad is, is it entirely based on ARM, and Apple has excluded an easy convergence of iOS and OS X in the beginning; though support for x86 does exist in Android, the Android tablet app ecosystem is still weak at best.

Of course, this world is built upon one assumption: Intel wins the mobile SoC war.

Whether this will truly come to pass, I believe with Windows RT and Windows 8, Microsoft is betting its future on the victory of Intel.