The Windows 8 demo from the All Things Digital conference left me kind of confused. More accurately, it left me thinking Microsoft is kind of confused. Perhaps most important, it left me thinking that most end users who pick a Windows 8 device are likely to be confused.
Don't get me wrong -- despite the fact that I primarily write about and work with Apple technology, I think Microsoft has a lot of good ideas going on in Windows 8. (It's worth noting that I said the same thing last fall about Windows Phone 7 as a potential iPhone rival.)
Let's recap the really good ideas that we've seen:
* A new lock screen that displays basic information for the user, à la WP7.
* A new live tile-based start screen (which in December I suggested should be part of Microsoft's tablet strategy).
* Simple tools for application development -- Microsoft took an excellent page from webOS here in allowing HTML5 to be a development solution.
* Unique gestures that haven't been implemented on any mobile platform thus far, particularly the snap gestures.
* Easy access to the file system and network resources from the new touch-centric interface; I give Microsoft huge props here over Apple's file management, or lack thereof, in iOS.
* New apps optimized for touch.
All of those are bold new initiatives for Microsoft in much the same way that Windows Phone 7 was. They're a complete rethinking and redesign of the way you interact with Windows. They make Windows 8 ideal as a tablet/mobile OS and could allow for crossover with Windows Phone 7 apps as well as the interface.
The fact that Windows 8's touch user interface borrows from Windows Phone 7 is very similar to the way Apple built the iPad's UI around the iPhones. It's also not unlike what Apple is doing with Mac OS X Lion in terms of bringing design elements and user experience lessons from a mobile OS to the desktop.
If I were to quibble about anything in the new touch-centric UI itself, I'd say there are two areas where Microsoft may be overreaching.
Too much gesturing?
First, the interface is gesture-intensive. By definition, all touch-based operating systems are -- to some extent. The problem is that Microsoft seems to be introducing too much gesture-only control. In addition to the usual tap, scroll, swipe and pinch that we've all become used to -- and which even toddlers can learn and master in minutes -- Microsoft is adding special gestures for system tasks.
The yet-unnamed control bar requires a swipe on the right side of the screen and appears to be the only way to exit an app, image or anything. Swiping on the left seems to be the core way to switch between apps. It's also how you use the new snap feature; which feature occurs depends on how you swipe. Then there's the vertical swipe gesture in Internet Explorer 10 to see the address bar. An upward swipe is needed to unlock a device.
That's a lot of context-specific gestures. While those kinds of movements aren't unheard of in other mobile platforms, they're usually limited and somewhat universal. To use an iPad or Android phone for the first time, you don't need to pick up anything but the very basics. WebOS, on the other hand, requires context-specific gestures and, quite frankly, is confusing if you don't know them.
This isn't a deal-breaker, but I think new Windows 8 users might find this confusing, particularly if there are more context-specific gestures than we've seen, which wouldn't surprise me. I think all of these will be incredibly non-intuitive with a mouse, but we'll get to that in a minute.
For what it's worth, Apple has built a lot of advanced gesture support into Mac OS X with its Magic Mouse, Magic Trackpad and notebook trackpads, and guess what: I use almost none of them, because it's usually too much effort to remember they're all there. Simplicity in gesture use is probably best for almost everyone.
The app question
The other issue I think Microsoft has to contend with is ensuring a selection of touch-centric apps designed to run in the interface. No matter how great the new interface is, if there isn't a robust selection of applications designed for it, no one will use it. Sure, you can run your old Windows apps (if you're not using an Intel machine), but they won't be touch-optimized and will run in a standard Windows 7 interface. That defeats the purpose of this great new touch setup.
Microsoft probably doesn't have a lot to worry about, since any Web developer should be able to create apps, but it will have to stay on the ball here. As I said, if the company makes it easy for Windows Phone apps to run (and scale visually) on Windows 8, it can easily have a viable selection. If the company works with existing WP7 developers to make an easy transition to the larger interface, it should avoid any problems.
The app issue will really impact Windows 8 on tablets more than traditional PCs because they are touch devices. If there isn't a diverse ecosystem, Windows 8 tablets will be stuck somewhere between Windows 7 tablets and Android tablets, where there is a dearth of Honeycomb apps compared to the iPad.
Touch on desktops and desktop-style tablet apps
The new touch interface is great for touch-enabled devices, be they tablets, hybrids or touch-screen notebooks, but it will seem stupid on traditional keyboard/mouse machines. Yes, it will work with a mouse, but many gestures will seem completely foreign to most users. And clicking the large tiles and other elements will feel like the interface was designed as a kindergarten computer.
This is going be problematic, since Microsoft is planning to make this UI the default on all devices. No matter how easy it may be to use this interface to launch traditional desktop-based apps or even to switch to just a Windows 7-style desktop, a lot of users are going to be confused -- particularly those who aren't that technically literate.
In fact, those same users are probably going to get confused when it comes to navigating the file system from both the new touch-based perspective and the traditional desktop interface. Even I had trouble keeping track of the differences being demonstrated during the demo, and I spend my life immersed in technology.
Then there's the opposite problem -- the traditional Windows desktop and existing applications in a touch-only environment. Microsoft has been pitching that paradigm for a decade, and it has never caught on or worked well. Now, not only will you need to revert back to it for some tasks, but it will be a jarring transition to switch from one to the other.
That's a big issue. It's the reason why Apple didn't try to shove any part of the Mac interface into the iPad or its apps. In fact, the same could be said about Android, webOS and RIM's PlayBook. They all have a specific UI design for touch, and they only support touch-centric and touch-oriented apps.
And that brings me to the real big problem with Windows 8: Microsoft has plenty of really great and innovative ideas here, but it's trying to push them, plus backward compatibility, together. The result is a mishmash of tablet/touch concepts combined with desktop computing needs that will result in a product that doesn't really excel at either.
Quite frankly, that's a real shame.
What Microsoft should do
Some people may argue that I'm bashing Microsoft and saying it should have acted like Apple. They'd be wrong. I think the new touch interface is innovative, powerful, unique and packed with potential -- but as a tablet OS, not as a tablet/desktop combo OS. A better strategy wouldn't have been to mimic Apple. Microsoft should be developing three flavors of Windows 8 based around what it's created:
* One version similar to Windows 7, but with support for running new HTML5-based apps while still offering the core desktop experience -- taskbar, Windows Explorer, and a version of IE that isn't touch-specific.
* One based solely around the new touch interface for tablets that eliminates the desktop while offering a way for developers to port existing apps to a more touch-based experience.
* And one similar to what it has chosen to do for touch-first and hybrid devices.
There is actually a precedent in Apple's playbook for a similar transition, but it's much older than the iPad: the transition from Mac OS 9 to Mac OS X. That's where Apple allowed older apps to run if needed but encouraged developers to move to a variation that ran native in both OSs, with an eventual move toward Mac OS X-only apps.
Of course, this is ultimately a lesson in slowly but surely ending backward compatibility and moving forward -- something that has long been an Achilles' heel for Windows.
Ryan Faas is a freelance writer and technology consultant specializing in Mac and multiplatform network issues. He has been a Computerworld columnist since 2003 and is a frequent contributor to Peachpit.com. Faas is also the author of iPhone for Work (Apress, 2009). You can find out more about him at RyanFaas.com and follow him on Twitter (@ryanfaas).