By Harry McCracken | Thursday, June 2, 2011 at 12:37 am
Analyst John Pescatore: “Other thank cloud computing, what’s the riskiest bet you’re currently making?”
Steve Ballmer: “The next release of Windows.”
–exchange at Gartner conference, October 2010
Looks like Ballmer wasn’t just blustering. “Windows 8,” or whatever it ends up being called, has a radically new interface–a sleek, touch-centric look that draws more on Windows Phone 7 and general trends in phone and tablet design than it does on a quarter-century of Windows history. Anyone writing about the operating system at this point needs to insert a disclaimer that we’ve only seen bits and pieces of it in action for a few minutes; that’s way too little to come to any firm conclusions pro or con. But we do know that Microsoft is going to attempt something big here.
In my post yesterday evening, I said that Windows 8 looks like the most radical change in Windows’ interface since Windows 3.0. It’s possible that that’s understating matters. By providing both the new interface and apps to go with it, plus the old interface and apps, Microsoft is asking users to live in two worlds in a way it’s never done before.
Except it has. This situation sounds a lot like the computing lifestyle that PC users lived with from 1990-1995 or thereabouts, when the commonplace state of affairs was to run Windows 3.x on top of DOS.
Back then, DOS was tried-and-true, and Windows 3.x was a reaction–albeit a rather delayed one–to the trends Apple put into place with the original Mac. Every Windows user ran it on top of DOS, since Windows was at that point an environment rather than a self-contained operating system. And the vast majority of users split their time between new-wave Windows apps and old-school DOS ones.
The two interfaces couldn’t have been much more different: DOS with its profusion of text, minimum of graphics, and command-line interface, and Windows with its graphical look and introduction of a still largely-unfamiliar input device called a mouse. Yet people learned to live with both of them. They’d run DOS apps within a Windows window. They’d exit out to a command line. They’d just not launch Windows at all if they didn’t need it at the moment.
They also had to continue dealing with the limitations of DOS, some of which creeped into Windows–such as memory-management hassles and short file names. Even so, people preferred it to IBM’s OS/2, a theoretically superior product that demanded that people take a great leap forward rather than allowing them to make baby steps away from DOS.
People also managed to deal with the fact that the typical PC of the early 1990s wasn’t built for Windows–small monochrome monitors were still common, RAM and hard-disk space were tight, mice were usually an optional accessory, and there was really no good way to run Windows on a laptop. But even Windows on a machine that ran Windows poorly was broadly appealing.
At first, even Windows enthusiasts spent an awful lot of their time in DOS apps, since the biggies of the day–WordPerfect, 1-2-3, Harvard Graphics, and other blockbusters–weren’t available in Windows versions. Over time, that changed. (Although it changed in such a slow, bumpy fashion that Microsoft managed to supplant most of the big names of the DOS era with its own Office apps.) Eventually, most people were left with only one or two DOS apps in their toolbox. The last two I remember running were ones I still miss: the DOS versions of InfoSelect and Norton Utilities. And by oh, 1997 or so, anyone who was running any DOS apps at all was a flaming luddite.
Daring Fireball’s John Gruber makes a point that seems self-evidently true to me: the iPad approach–a total replacement of the Mac user interface with a modern one built purely for touch–is conceptually superior to bolting a new touch interface on top of an old PC platform. It’s elegant. It’s decisive. It doesn’t burden a platform of the future with all the baggage of the past. It’s also parallel in some respects to what Apple did with the original Macintosh, when it chose not to maintain any Apple II compatibility whatsoever. By contrast, Gruber says, Windows 8 is “fundamentally flawed.”
(Apple did try to build a Mac-like experience on top of Apple II technology eventually: it was called the Apple II GS, and the world rightly identified it as a misguided idea and ignored it.)
Adding a new touch interface to old Windows isn’t elegant or decisive, and it does burden a platform of the future with all the baggage of the past. But if Microsoft manages to pull off a transition this time around that bears any resemblance to the DOS-to-Windows transition, its strategy could be “fundamentally flawed” and wildly successful. I’m not saying it’s going to happen. But it’s a best-case scenario for Microsoft that doesn’t feel completely wacky.