Why 10/GUI is brilliant and will probably never work
10/GUI is a multitouch interface designed to push the typical desktop experience forward with the “interaction bandwidth” afforded through the use of all ten fingers. It’s smart, inventive and really inspiring.
While Microsoft’s Courier concept is fascinating for its application of touch to a decidedly alternative computing experience, 10/GUI seeks to redefine our desktop interactions. I was pleased to see its creator, R. Clayton Miller, thoughtfully address the issue of arm and neck fatigue (a problem cartoonists like myself know all too well). His solution also counters the challenges of the user’s fingers obscuring the point of interaction, something today’s mobile devices sidestep with clumsy fly-up keypress confirmations.
I think 10/GUI is wonderful, and I sincerely hope Miller (or those industrious enough to seek him out) will explore it further and give us some real products to play with. That being said, I remain unconvinced that this type of interface would work in mainstream application.
While listening to 10/GUI’s daunting list of touch gestures required to accomplish basic operating system tasks, I was reminded of industrial designer Dennis Boyle’s account of Palm’s experience selling users on Graffiti text entry over a traditional keyboard. From Bill Moggridge’s Designing Interactions:
I remember that Handspring decided to put the Treo out with both a keyboard and Graffiti, because they didn’t know which one people would choose; they decided to let them vote. The result was quite clear; a large majority went for the small keyboard. [...] [The] little QWERTY keyboard, bad as it is, is such a standard that it requires no guesswork, and that attracts more users.
Despite the fact that Graffiti was a faster and more efficient method of “typing” on a mobile device, these rewards came only if the user invested enough time to overcome the obstacle of learning the standard. While I find it to be really neat, I can’t imagine gaining enough speed from 10/GUI to overcome the time and pain it would take to learn those gestures.
I’ve experienced this with my MacBook Pro’s multitouch trackpad. While simple gestures like scrolling are easy to learn, more complex maneuvers are bothersome and often unwittingly triggered as my hand brushes past while typing. While some of these features ease the frustration of not having a mouse, they are a poor substitute when used in conjunction with the keyboard, occasionally even to the detriment of my workflow.
I’m not saying we necessarily need to coddle our users, gimping innovation for the sake of complacency. What I am suggesting is that immediacy may be the greatest asset of touch computing. It is natural to select an object by touching it. I love the Courier prototype because it evolves tasks which inherently benefit from touch interaction. 10/GUI appears to require memorization prior to use, more like a musical instrument than a user interface. Put simply, I’m afraid 10/GUI will create more problems for me than it will solve.
But I really hope I’m proven wrong.