Top 10 Windows 7 Features #5: Multitouch

The immediate upside of multitouch What the engineers are learning from this process is already having a positive benefit to Windows 7 in general, in features you'll see and use every day. The concept is being called Aero Snap, but it doesn't actually require the Aero rendering model at all -- it works fine in a virtual machine without Aero. After 19 years of the current window model, Microsoft discovered something that could have helped them as far back as 1984, during the days of the "MS-DOS Executive:" Perhaps a simpler way to maximize a window would be to just drag its title bar to the top of the screen.

Now, while that concept sounds simple enough, what if all the user's trying to do is drag an existing window up? If he hits the top of the screen, its size could blow up unexpectedly -- and once again, Windows feels like a foreign substance. The user experience (UX) engineers tackled that problem by creating a way to "feel" when a window dragged up there can be minimized. If the left button is held down while the pointer touches the edge of the screen and stays there, Windows 7 responds with a visual signal that highlights the entire workspace, advising the user that the window would be maximized if he let go of the button. (This effect does look cooler in an Aero environment, where the system applies a glossy finish; in normal rendering, the system merely covers the workspace in a blue halftone.) If the pointer continues beyond the upper edge, the window is merely dragged on up, just like before.

Taking this idea a few steps forward, the Win7 user can now drag a window to the left edge to have it "semi-maximized" -- to fill just the left half of the screen. The visual cue remains, so if the pointer moves to the edge and stops, he's given a clear warning with the visual effect. But then the user can do the same with the right edge as well, providing at long last, after two decades of complaints from me and Jerry Pournelle and the rest of the world, a way to create a dual-paned Explorer environment in about a second. And having learned lessons from the Surface and Touch projects, Win7's UX engineers remembered that doing the opposite motion should provide the opposite response; so once snapped windows are dragged away from their edges, they return to their previous size and shape.

This is the beginning of something, a lesson finally learned after years of plodding in the same general direction with the same general result. The introduction of styluses and now fingertips into the input model has finally led to a scenario where apps may at last be able to process natural language syntax -- not just single words like "maximize," but sophisticated concepts like, "Show me what that might look like in five years." Simply adding the input analysis phase as a step in the process of computing will -- if we do this right -- truly revolutionize the way we work. Of course, if we follow the same patterns we've followed up to now, the whole idea could also begin and end with Snap.