The Windows 8 tragedy: How Microsoft can avoid disaster

Microsoft’s Windows 8 beta has been publicly available for nine days, and it’s clear hardly anyone likes it. IDC predicts Windows 8 will be “largely irrelevant for users of traditional PCs.” Tablet sales, it says, will “disappoint.” I’ve talked to more than a dozen people in IT and the user community, and I’ve found no fans. The reaction is a sure sign that Windows 7 will remain the desktop OS on the dwindling pool of traditional PCs for years to come, but the new world of devices — tablets, smartphones, and ultralights, both stand-alone and dockable — will run something else: increasingly the converging iOS and Mac OS X, and perhaps Android.

Windows 8 is a real tragedy, because unlike the Windows Vista debacle, Microsoft actually had a clue that the world is changing and is trying to adapt to it. The time has come to retire the Windows as we’ve known it since Windows 95. That’s what Microsoft is essentially trying to do with Windows 8, making the new Metro UI (or Metro OS, on some devices) the default “OS” for users and relegating Windows 7 as the legacy OS behind the curtain.

I know that upsets many Windows users, but it needs to be done. It’s too late for a gradual transition, as Apple has been undertaking for five years with its iOS reset and measured Mac OS X convergence. But sadly, Microsoft’s execution on this necessary but difficult transition is threatening the whole effort, and it’ll likely hasten both Windows’ and Metro’s retirement.

Microsoft can fix some of the mess it has made so far with Windows 8 — before the product ships this fall and falters in the face of iOS, Mac OS X, and even Android. I hope it does, so I’m offering my suggestions in this post. We need a strong competitor to Apple.

Why people don’t like Windows 8

Several commentators have likened Windows versions to “Star Trek” movies: Only every other version was good. In that view, Windows 7, XP, and Windows 95/98 were good, and Windows Vista, 2000, and Me were bad. It’s not quite every other one, but close enough. The issue is not uneven execution — though there’s some of that. The basic problem with Windows 8 is that it forces together two very different OSes — Windows 7 and Metro — in an unnatural combination and is applied monolithically to several kinds of device contexts.

The InfoWorld Test Center’s excellent first-look review of Windows 8 details the confusing and brain-sapping relationship between the traditional Windows 7 desktop and the new Metro UI or, more precisely, the lack of relationship. My colleague J. Peter Bruzzese — InfoWorld’s resident Windows admin expert, a true Microsoft fan, and MVP-certified Windows admin — has been brought close to tears by Windows 8’s dual personae. My colleague Woody Leonhard — a longtime Windows expert — is also befuddled by the parallel universes that inhabit Windows 8. InfoWorld’s Neil McAllister also flays Metro’s poor UI and exposes Microsoft’s incoherent approach to Metro’s usage.

What Microsoft has done is shockingly stupid. Windows 7 is a very nice OS, but it’s unusable in a touch environment. Yet Microsoft is promoting touch as the main user interface method on all Windows devices, from tablets to traditional PCs and laptops. It simply doesn’t work, so to use the Windows 7 part of Windows 8 — and Windows 7 is required for most of your traditional applications — means you must have a keyboard and mouse. That makes Windows 8 on tablets an iffy proposition. If you have to add a keyboard and mouse to a tablet, you essentially have an assemble-it-yourself laptop with multiple pieces to misplace.

The stupidity doesn’t stop there. The Metro UI is very touch-oriented, but it works poorly with a keyboard or mouse. People will eventually grow accustomed to using these input devices with Metro, but doing so requires relearning how to work with these devices — and remembering to switch back to the old methods in Windows 7. Even if you have a touchscreen-equipped PC, where you might use touch for Metro and a keyboard and mouse for Windows 7, you’re asking for users for way too much fundamental context-shifting.

Then there are the idiotic decisions in basic UI areas. For example, the Start menu is gone from Windows 7. That button is the central starting point for all of Windows, and now it’s gone. Yes, you can use keyboard shortcuts to access its functions, but why? It’s a bizarre throwback to a DOS mentality and not very helpful for touch users.

In Metro, the three menu bar areas (charms at the right, contextual options at the bottom, and settings and other functions at the top) are simply confusing. Why not one menu area to focus on? After all, that would greatly simplify touch selection. You’ll also encounter unnecessary frustrations, like getting back an app you accidentally unpinned from the Start screen. I’m sure it’s possible, but I haven’t yet figured out how. Given how easy it is to unpin an app, it should be just as easy to repin it.

Then there’s the confusion over Windows 8 on x86/x64 processors (WOX) and Windows 8 on ARM (WOA). WOA has only Metro, whereas WOX has both Windows 7 and Metro. That’s related to the chip architectures, but users will be confused nonetheless. WOA doesn’t support basic Windows management such as domain joining, forcing it to be handled differently by IT — as if it were an iPad or Android, managed through a mobile device management (MDM) tool rather than as any other Windows client. That’s just dumb — and Microsoft should be able to figure out how to port the client portion of its admin capabilities to WOA.

These issues are different in nature than those in the mess that was Vista, where the two biggest complaints were a user interface that exploded with disorder and confusion, and a security model sure to frustrate users to the point of making it insecure. The basic problem with Windows 8 is that it has two essentially independent UIs: a traditional one for PCs and apps, and a new one for smartphonelike devices and apps. Yet you have to switch between them regardless of the type of device you are using.

Compare Windows 8’s Jekyll-and-Hyde approach to the smooth transition Mac OS X and iOS users experience as they move from one device to the next. There’s a commonality between the two, but desktop UI is not imposed on mobile devices, and mobile UI is not imposed on desktop devices. Apple has gradually trained Mac OS X users how to use touch gestures by supplementing Mac OS X Lion and the forthcoming OS X Mountain Lion with gestures, but it doesn’t require their use and thus doesn’t force the kind of awkward user experience that Windows 8 does.

What Microsoft needs to do to save Windows 8

Microsoft is way late to the new world of mobile, touch-savvy, hetereogeneous devices, so it’s clearly trying to force users to adapt in one fell swoop. That’s why you must go through Metro to get to Windows 7 in Windows 8. But Microsoft’s clumsy approach is likely to create a backlash against such adaptation. Worse for Microsoft, it could open people’s eyes to the new world and lead them to Apple’s and, for mobile devices, Google’s more intuitive, mature user experiences.

Remember: With Vista, Microsoft made the world realize that life goes on just fine if you skip a new Windows version. With Metro, Microsoft has endorsed the vision that Apple created for the post-PC world, and that means the Apple option is as valid as Microsoft’s version of it. Microsoft only wins in the new world order if it does at least as good as Apple; it can’t afford more “every other version sucks” execution — especially not in its transitional product for the new world.

Here’s what Microsoft needs to do:

Separate Metro from Windows 7. Metro is the future of Window, and Windows 7 is the final version of the legacy Windows. Microsoft should offer Windows 7 Plus and Windows 8 Metro as two distinct products, even if they come in the same box or download as “Windows 8.”

Companies with homegrown apps, as well as a subset of users that work with Excel’s extensive macros or run core ERP apps, will need Windows 7 for years to come — so keep selling Windows 7 for such legacy applications, legacy PCs, and legacy purposes. Microsoft claims the Windows 7 environment in Windows 8 has had its code size reduced and its performance optimized. If that’s true, great — make it an automatic update for current Windows 7 users, as well as part of Windows Plus. Ditto with other enhancements to Windows 7, such as its improved system utilities and support for iCloud-like fabric computing via Windows Live.

Windows 8 should be smart enough to make Windows 7 the default OS for nontouch computers and devices. I can see training users on Metro by making Metro available as an app in Windows 7, sort of like Microsoft once tried for its widgets — and as Apple does in Mac OS X for its widgets. But Metro should not be the gate through which traditional PCs get to Windows 7, and Windows 7 should keep its Start menu; it’s too useful to die.

On tablets, Metro should be the default OS, and Windows 7 should remain hard to access. Windows 7 is a last-resort legacy environment for those apps you must use but haven’t been modernized — similar to working with a VDI client on an iPad to access legacy apps. It’s the tablet where Metro makes the most sense and where Microsoft should be training users on the “new world” OS that Metro is.

As for touch laptops and PCs, give users the ability to set the default gate. Chances are most of their work will be traditional Windows duties, so they’ll want the regular Windows 7 by default. In Windows 7 jobs, they won’t be using the touchscreen — nor should they. But if they want to switch to Metro for its touch-savvy apps, let them.

In my recommended approaches, the OS that makes the most sense for the current device is what users experience first and foremost. And the “other” OS is not in the way.

Rework Windows 7 to be touch-proficient. This won’t happen by the expected fall release of Windows 8, but if Microsoft really believes that Windows 7 should live beyond legacy, it needs to make that OS touch-savvy — and not that crappy “Windows for Pen” environment it’s been trying to foist on users since Windows XP. Touch means using fingers, not pens. The truth is that, on touch devices, Windows 7 is a bad UI — the icons, menus, and other UI elements are too small to touch.

The place to start is to have contextual DLLs for the major UI components that most apps use; the touch-savvy UI appears on a touch device by default, and users can switch the display mode if they’re using a mouse or trackpad. Windows 7 and its applications should be at least usable via touch gestures if they’re to run on touchscreen devices.

I realize this approach is clumsy and will mean work by app developers, but it’s more elegant than the poor experience Windows 7 now offers. Of course, the real answer is to abandon Windows 7 as soon as possible. Realistically, that will take years, and if Microsoft insists on carrying Windows 7 into the modern world, it needs to modernize Windows 7 enough to work plausibly. Apple has done a good job of this in Mac OS X Lion and OS X Lion, so use that transition as a guide.

Get rid of Metro’s x86/ARM distinction. It’s nuts that WOA isn’t equivalent to WOX. The reason no doubt is that WOX relies in legacy Windows for native Windows client management, signaling a failure at Microsoft in its effort to port that technology to the ARM processor. WOA shouldn’t ship as a second-class Windows Metro. In fact, users should neither know nor care whether their Metro tablet uses an x86 CPU (from Intel or AMD) or an ARM CPU. Neither should IT. That means WOX and WOA need to have real apps and be equal in all respects.

If users must lug around Windows 7 on their tablets “just in case,” they can buy x86 tablets that have both Metro and Windows 7 Plus. Likewise, if they buy into the notion of convertibles — laptops whose screens detach to become stand-alone tablets — they’ll get a WOX device. But a pure Metro tablet should have the same capabilities regardless of the CPU architecture.

Make Metro a strong OS in its own right. Microsoft seems to be treating Metro as a UI overlay to Windows 7 on all but ARM devices. That makes it too easy to rely on Windows 7 as a crutch that lets Metro stay underpowered. Imagine if Apple had put Mac OS X on the iPad, as many pundits demanded: App developers wouldn’t have bothered creating compelling native iOS apps. But Apple forced them to do so by providing no such crutch. The fact that all Windows devices but ARM-based ones can run Windows 7 (albeit poorly) means that Windows developers will be slow to take advantage of Metro — and will simply perpetuate the Windows legacy.

Even if Microsoft separates the two OSes as I suggest, that Windows 7 legacy crutch remains and will slow the development of strong native Metro apps. To combat that laziness, Microsoft needs to make Office 15 for Metro a truly compelling app. It needs to be at least as capable as Quickoffice or iWork on an iPad. To do less is to proclaim that Metro tablets are fancy e-readers — niche products like the Amazon.com Kindle Fire, not true computing platforms like the iPad.

But even an equivalent to Quickoffice or iWork is insuffficient: Microsoft needs to make Office 15 on Metro uniquely compelling. In other words, it needs to signal that Office for Windows 7 is done, and the innovation will be only on Metro. As Apple’s work on iWork, iMovie, and GarageBand for iPad set both the example and the bar for iPad developers, so must Microsoft do for Metro. That will be hard for Microsoft, which is too cozy to its enterprise customers’ inward focus and conservatism to tell them it’s time to make the change. Orphan Windows 7 where it is, and put the future in Metro. They’ll complain but adapt — otherwise, their only other adoption is to go to Apple, which they know will not indulge their desire to stick with the old forever.

And Microsoft needs to give Metro much more capability and sophistication. The UI may be clean, but it’s very limited. iOS and Android can run circles around Metro in so many ways, it’s pathetic. Metro is an OS to run widgets, not a serious computing platform. Microsoft needs to change that soon, or else Metro will be just as irrelevant on PCs and tablets as it is in smartphones.

Is Microsoft able to bet right on its own future?

If you listen to Microsoft’s marketing (you shouldn’t), you’ll hear that Windows 8 is the culmination of Manhattan Project-like usability research scientifically proven to be the one true approach for humanity. It’s not. The Metro UI is a blownup version of what Windows Phone 7 offers: a clean, simple interface for running a handful of widgets. It works poorly on a large touchscreen, and it works terribly with a keyboard and mouse. Windows 7 is likewise completely inappropriate for the touch environment. These UIs would not have survived real usability research.

Every time Microsoft has a new OS, it changes the UI. If the UI were based on legitimate usability testing, it wouldn’t need such radical changes every few years. That constant level of significant UI change simply means Microsoft is still getting it wrong. Compare Windows to Mac OS X, which had a major retooling 12 years ago. It has evolved dramatically since, but no one has ever had to relearn the Mac UI from version to version. A Mac user from 20 years ago would recognize the fundamentals in today’s version, as different as they may appear on screen. That should have been the history of the Windows UI as well.

I saw a Microsoft presentation this week on Windows 8 and all its glories. The marketing exec painted a vision of thin laptops and tablets as what Windows 8 would make possible. As I looked around the conference, I saw dozens of MacBook Airs, scores and scores of iPads, and scads of iPhones and Androids. I also saw a lot of rolling eyes as the Microsoft rep continued to embarrass herself with clueless claims. That future Microsoft pretends it’s inventing already exists, and today’s reality is far ahead of Microsoft’s “vision.”

The foks in Redmond need to stop telling each other how wonderful they are. The emperor has few clothes left.

I can’t understate the danger around Windows 8 as it exists today. The current implementation will fail, and that will be worse for Microsoft than its previous debacle, Windows Vista.

Vista was so bad that more than 210,000 people signed InfoWorld’s successful petition to keep XP alive. Thanks to Vista, Windows XP remains the primary desktop OS into its 11th year of existence, whereas Vista never had more than 20 percent of the market, even though nonenterprise buyers had no choice but to have it preinstalled on new PCs. Also thanks to Vista, we’ve all learned we don’t need to install a new version of Windows just because Microsoft has one. Users and IT agonized about not adopting Vista — I hear no agony this time in either camp about skipping Windows 8.

Worse, Microsoft has enough of a clue in its Windows 8 efforts to fool itself that it’s doing the right thing. Vista was a disaster, and they knew that privately within Microsoft. When the marketing campaign couldn’t hide the stink, Microsoft fired the culprits and Steve Sinofsky came in to clean up the mess, giving us Windows 7. The same pattern seems to be finally occurring in the mobile space, after the Windows Mobile 6, Kin, Windows Phone 7, and Windows Phone 7.5 failures: The new hope is Windows Phone 8 “Apollo,” ostensibly led by Sinofsky’s people.

That makes it easier to deny the problems, because there’s just enough success to hang on to. Although Windows fans such as InfoWorld’s Bruzzese and ZDnet’s Mary Jo Foley are plaintively warning Microsoft about the train wreck coming, some Microsoft fanboys such as Paul Thurrott have told their readers to basically suck it up and adapt. Certainly, change requires adaptation, but why should anyone adapt to a mess? The last time Microsoft gave us a confusing OS, people adapted — to Mac OS X. Now they’ll adapt to iOS and Android as well. Users don’t have to choose Windows, and if Microsoft is going to make us all relearn how to work with a PC, we can now decide to spend that effort on an item of our choosing. Adaptation should be worth the effort.

Users have seen the iPhone and the iPad, and they’ve realized there were other ways to compute. The growth in computing is now all in iOS and Android smartphones and tablets, not in traditional PCs. Mac sales are also growing, and you know a fundamental change has occurred when you see consultants from old-guard IT consultancies such as CSC using MacBook Airs at conferences. It’s as if users discovered California cuisine after a steady diet of Wonder bread and peanut butter.

Where Windows 8 has gone wrong is that Microsoft correctly realized that the traditional PC paradigm of mass-produced boxes that make users adapt to them is dying, but is mishandling its shift to the new paradigm. The market now truly knows better — that wasn’t necessarily the case during the Vista debacle.

I’ve said several times that Windows 8 has a shot at bringing Microsoft into the new world, providing a strong alternative to the Mac OS X/iOS juggernaut. I meant it. But Windows 8’s current form and trajectory will blow that shot, and time is running out.