Windows 8’s GUI: A Tale of Three Desktops

You’ve probably heard by now that Microsoft is about to release a beta of the next version of Windows, code-named Windows 8. (Ah, for the days of code names like Starfighter, Memphis, and Worf!)

As I write this, Microsoft has revealed only a handful of facts about Windows 8, one of which I find pretty interesting: Windows 8 will have three different desktops. (I know, you might have read that it’ll have two desktops, but that’s not really correct, as I’ll explain in a minute.) The new desktop looks a lot like the Windows 7 Phone interface, which basically shows applications not in the way that we’ve seen in Windows or the Mac over the years, but instead as “tiles”—rectangular things that are sort of halfway between an overgrown icon and a standard Windows-type window but without the frame, menu, and so on. (In case you didn’t know, a window without any of that stuff is said to be “chromeless”—I mention this because I’m pretty sure we’ll be hearing that word a lot when Windows 8 comes out. The word’s been around for a while, but it hasn’t been used much up to this point.) Apparently developers use a different set of tools to build these new sorts of Windows apps, which Microsoft has dubbed “immersive” applications, and those applications run a completely new desktop, the “immersive” desktop. And here’s the scary part: Apparently a Win 8 box will show either the “legacy desktop” (my phrase) or the immersive desktop at any moment in time, which would mean that when you switch from, say, Lightroom 3 (which was built for the legacy desktop) to Imaginary Immersive Office 2012 (built for the immersive desktop), then the whole computer screen changes from the legacy desktop to the immersive one, and you can’t see any of your legacy apps until you shift back to the legacy desktop.

Now, don’t get me wrong; I’m not intending to seem aversive to the immersive—I’m totally addicted to my iPad and would love to see a tablet based on my favorite OS that’s more powerful, better-looking, and more open than iPenitentiary, in which Apple incarcerated my iPad. Clearly the immersive UI is built for tablets, and it might really do the job for that market segment. I do, however, foresee a few problems and have a few questions.

First of all, isn’t this desktop-shuffle going to confuse most people? I mean, think about it: There are plenty of folks still running Windows 98, for heaven’s sake, because all they need is a web browser and email. How many lawsuits will Microsoft see from DSIMS (desktop switch-induced motion sickness)? And how hard (or, preferably, how easy) will it be to move data between apps on the two desktops? I’ve grown fond of drag-and-drop over the years, and I’d hate to lose it.

On a personal note, I must admit, I’m kind of looking forward to this three-desktop world. I mean, at first, I was troubled by the desktop shuffle thing, because I’d hate to see my favorite OS suffer another market failure, but then I finally remembered something very important: I make a living explaining mysterious things about Windows. That’s precisely how I pay the rent. So, the more complicated and/or goofy a new version of Windows is, the more business I can anticipate. And at that moment, Windows division president Steven Sinofsky became my new best friend, even though I’ve never met the man.

Second, what does this mean for the legacy desktop’s future? Will we see Win32/.NET-style applications fade away? My initial guess is no, because many of us have spent bazillions of person-hours learning to write that kind of stuff. The new immersive apps are based on HTML5 and JavaScript (according to what I hope was an unwise and incorrect comment by a Microsoft rep), and it’s kinda hard to imagine rebuilding Office, Lightroom, and AutoCAD with those, um, almost-programming tools. (Really, Microsoft? Javascript? I mean, if it wasn’t going to be PowerShell, couldn’t it at least have been VBScript, the scripting language known by more Windows techies than any other?) So who knows, maybe the new desktop will be known as the “fluffy desktop” and the legacy desktop will be seen as the “getting things done desktop,” as in, “Hey, do I run Acrobat on the real desktop or the tile one?”

Adding a desktop is also an interesting thing (to me, anyway) because Windows has actually done this once before. Every version of Windows up to and including Windows 7 has not one but two desktops: the GUI desktop and the original desktop—the command window, or what might be called the “DOS desktop.” That desktop enjoys a significant, large, and powerful set of tools without which many administrators’ lives would be greatly complicated (i.e., the myriad of command-line tools that exist for Windows). So, between 1990 (Windows 3’s release date) and now, what’s happened with the command window? Has it faded away?

Heck, no! In the years since Windows 3 arrived, we’ve gotten the ability to window thousands of lines, we’ve gotten cut and paste, and the batch language that first appeared in DOS 1.0 has grown tremendously. (For a couple of examples, consider the text parsing abilities in the FOR command and the substring/arithmetic capabilities in the SET command, both of which appeared at the turn of the century.) And, oh, did I forget to mention, the past few years have seen Windows get a command-line shell that quite frankly kicks the butts of all of the UNIX/Linux shell scripting languages, an honest-to-goodness object-oriented command-line platform? (Thank you, Mr. Snover.) In short, I think that history has shown that maybe—just maybe—Windows can survive Multiple Desktop Syndrome, but we’ll see. And, that left-handed reference to PowerShell leads to Question Three.

Third, will immersive apps be automatable? When Microsoft went big with its GUIs, the company was hoping to win the world from the UNIX-ish OSs, which were extremely rock-solid but difficult to administer, with an OS that was so-so in the reliability department but pretty easy to run, assuming you had a graphics card and a mouse. In the process, Microsoft appealed to the less-technical user base, which in other words was most of the world. Unfortunately, however, graphical applications have tended to be difficult to automate, because you can’t put mouse clicks in a batch file, and the ability to automate things in the Windows world has become a necessity in modern enterprises. With a new desktop comes a new opportunity to do it right from the beginning, so I hope that the new Windows desktop makes it easy for developers to build automation into their immersive apps.

So, in short, Microsoft, permit me to suggest that you make the two desktops play as well together as is possible, don’t tick off the legions of developers who already know and love Win32 and .NET, and make us admin types happy by ensuring that the new desktop is very automatable—and you might actually have a new interface that we’ll all want to immerse ourselves in.