Touchscreens are the rage these days, what with all the speculation about an Apple tablet. Nonetheless, they've got some heavy drawbacks: Given that they don't provide any tactile feedback, it's hard to imagine any laptop or desktop ever using them. Would you really want to spend hours a day tapping at something that's basically an overgrown iPhone? Not only is it not ergonomic, your hands are likely to obscure whatever it is you're working on.
The basic hardware would be a keyboard with a big multi-touch panel below it. Touching the panel's right edge brings up a "global menu" that decides what applications you launch; touching the left edge brings up a "local menu" for whatever application is in front of you. (See 5:19 in the video)
In addition, the number of fingers you use "determines the level at where your gestures have effect." Two fingers, you're controlling an application. Three fingers, and you can move your applications around. Four fingers scrolls through the applications. (See 6:00 in the video.)
There's one other idea in Miller's concept: Applications and windows are all forced into a single, organized deck. That means less freedom, but it also eliminates the clutter created by piles of windows--and makes the gestural grammar a bit easier to digest.
And that's really the key with such a sprawling system--and the main challenge in creating any new computer interface. How do you make it do completely new things while keeping it simple enough so it takes only one sit-down to master? While 10GUI looks fairly complex in the video because there are so many components to the interaction, don't judge it until you use it. That's the alchemy of interaction design--explanations and complexity disappear if the design is good enough. Just think: How weird was it when you first saw a video of the iPhone being used? And how easy is it to use now?
[Via Ignore the Code]