Why Windows Is Obsolete


startup repair

Windows 7 was rude enough to crash catastrophically Friday, an hour before a column deadline. After 45 minutes of hopeful auto-repair, an error message unceremoniously notified me that, no, nothing will be repaired automatically today. “Would you like to shut down?” Um, no, I wouldn’t.

The only solution was to reformat the disk and rebuild the system — a procedure I’ve done maybe 50 times since first installing Windows 3.1 in the early 90s.

I’m thankful for Carbonite. As of this writing Tuesday, the online backup servive has been restoring my 172 gigabytes of data since Friday. So far it’s 22 percent done. I should be all restored up by July.


The failure took my column with it, of course. I had to re-write it on my iPad.

Which got me thinking about the future of Windows, a future looks bleak from the perspective of turning to an incredibly stable $500 appliance while my $2,000 PC restores itself after yet another crash.

Don’t get me wrong. I actually think Windows has always been underappreciated. It’s an OS designed to run the widest possible range of software on the widest possible range of hardware. It runs 15-year-old shareware just as easily as it does brand-new enterprise management software. The company gets a lot of grief, but it achieves something every day that no other company has been able to achieve, which is to sell the world’s most widely used platform without actually controlling either the application software or the underlying hardware.

Microsoft’s model is actually breathtakingly ambitious. Unfortunately, it’s also obsolete.

In the 1990s, Windows was in fact the best model from a business perspective. Before social networking, the Web 2.0, cloud computing, apps, and all the rest, applications were the only game in town. What most users wanted was more of everything – more features, more applications, more capabilities. The Windows platform had far more applications in just about every category than its competitors in the market. In some categories, such as gaming, Windows was for years the only significant platform player. Microsoft’s model of supporting everything paid off, made the platform totally dominant, which enabled Microsoft to get into the habit of charging a lot for both its OS software, and also application suites.

But now the world has grown infinitely more complex. We no longer want more. Now we want less.

Two new alternative models emerged in 2007 for giving users less or charging users less that will inevitably destroy the Windows model — or at least relegate it to minority status.

The first is the “appliance model,” represented by Apple’s iOS. Apple has always tried to make appliances, but with moderate success. The iOS is a true appliance, where the hardware has been developed from the ground up for the software, and the software has been developed from the ground up for the hardware. The whole system was designed from the perspective of not “how do we make a PC better,” but “what should the user experience be?” Third-party application developers have been thoroughly fenced in, and are locked out without being approved by the governing company. The user experience for a true appliance, such as an iPad, is that it just works. There’s no managing anything, or maintaining anything. Users simply add or delete apps, and use them.

Apple is uniquely skillful at designing, building and selling such appliances, and will dominate the appliance market completely and indefinitely.

The second model is the “radically free and open model,” represented by Google’s Android. This model is closer to the Windows model, but it’s even more extreme — and a lot cheaper. It’s more extreme in that Google makes even fewer attempts to control the ecosystem. Anyone can grab the code and modify it to place it on any hardware they want. Not only can anyone write apps for Android, but they can even build and maintain their own app store.

These two models will decimate Windows’ market share in the coming decade.

I predict that by 2020, Google’s Android will have two thirds of global market share for all OSs, both mobile and desktop and Apple will earn two-thirds of all platform revenue. Microsoft will be number two in both market share and revenue.

To say that in another way, I believe that within nine years, the three biggest platforms from a number-of-users perspective worldwide will be Google, Microsoft and Apple in that order, and the biggest bucks for platforms will to go Apple, Microsoft and Google, in that order.

Google will dominate user numbers because Android is free and open. This will appeal to OEMs, developers and users in emerging economies like China, India and Brazil, as well as specific sectors of economies like governments, militaries and others.

Apple will dominate revenue because it will remain a luxury brand for people in rich countries. Right now, iPads are on the low side of the tablet pricing spectrum, but I don’t believe this will last. Within a year or two, Android tablets will drop to a couple hundred bucks. Apple will continue to sell iOS devices at a hefty profit, as well as profit massively on its cut of the app market.

Right now, the new models dominate mobile. But they’ll increasingly encroach on mainstream computing. It’s just a matter of time before Apple is selling desktop iOS appliances, and Google is selling desktop Android systems.

Alternative platforms from HP, RIM, MeeGo, Maemo, Bada and others will gravitate strongly to either the appliance or free-and-open model.

Poor Microsoft.

Only Microsoft will be left trying to compete in the market offering neither the ease-of-use of an appliance, nor the freedom and flexibility of the free-and-open model.

The new models are going to be brutal to compete against. Everyone who can afford Apple’s relatively pricey appliances will keep lining up to get them. Everyone who cares about price, customizability and flexibility will jump on the Android bandwagon.

Microsoft will be left to compete for scraps with the copycat also-rans, leaning heavily on its long-time corporate customers and experimenting with various models with limited success.

At least that’s how it looks from my perspective. (Oooh, look! Carbonite has now restored 23 percent of my files!)