ryan.freebern.org

What’s the ideal way to tell our computers what we want to do? Basically, we need to communicate two types of data: symbols (text, numbers, etc.) and kinetics (for lack of a better term–moving, clicking, dragging, etc.). Ideally, all we’d have to do is think about what we want done, and the computer would do it. Sadly, while progress is being made in the direction of a neural interface, I can’t imagine a general-purpose one succeeding in the market for at least another decade. So, at least for the time being, we need to focus on improving the current physical interface that allows us to use our hands to do the input.

The keyboard is a fantastic invention: an arrangement of buttons designed to let a person type any word or string of symbols they choose. However, the QWERTY layout (it is often said) was designed to slow typewriter typists down, so that the typebars wouldn’t jam. This is obviously not a concern with modern keyboards, but do we ever see any serious attempts to correct the layout and increase typing speed and productivity? Dvorak has sadly never caught on. And worrying about the layout doesn’t even touch on the fact that, with the addition of the ubiquitous computer mouse, the typical computer interface is optimal for a human being with three hands.

The mouse is a stopgap measure. Originally created to act as a sort of finger-proxy, a way to translate the intuitive human ability to point, push, drag, and tap with the fingers onto non-sensitive screens, it could be entirely superfluous now that touch-sensitive screens are both available and relatively inexpensive. But we’re stuck in our rut, and we don’t want to give up our three-handed interface, so we continue using a system that would be perfect for a mutant but requires wasted effort for most of us.

What to do?

Let’s face it: half-assed measures won’t cut it. If people have the option of sticking with the familiar suboptimal interface, they will. The redesigned computer shouldn’t give people this option. Its redesign must be so compelling, so obviously superior, that people will be willing to make the leap from the comfortable to the new. We need a powerful, intuitive new interface that eliminates the three-hand problem while still allowing us to do everything we already do and more.

Lots of research is being done into new input interfaces for PDAs and mobile phones and small-sized tablet PCs, but very little time and effort is being spent on regular old desktop or laptop input interfaces. I’ve spent hours looking into one-handed keyboards like the Matias Half Keyboard, the Frogpad, and the Twiddler, but they all suffer from one notable problem: typing anything other than regular English text is very difficult, so programmers like me will experience significant slowdown, which can be a stumbling block to staying in the creative zone. I’ve checked out on-screen writing software like Quikwriting and Dasher, but while these would be great for mobile phones or PDAs, they’re far too slow for regular use.

The makers of the Frogpad have begun mentioning that they’re working on a Frogpad for programmers that makes typing punctuation and symbols easier, but until that comes out, I think we’re effectively stuck with a two handed keyboard. So what can we do?

I recently took an informal poll of some friends to find out how they hit the spacebar on their keyboard while typing. Over 4/5ths of them answered that, like me, they always use the same thumb to type a space, leaving the other thumb always hanging there uselessly while they’re typing. An eye-tracking system uses cameras to track where on the screen you’re looking, and requires you to use something to input mouse clicks. Why not use that other, useless thumb? Instead of having a giant spacebar on your keyboard, just cut it down slightly and use the extra space as a “mouse key.” Then, to move the mouse pointer around the screen, you just have to look where you want it, and to click you just tap that key. Sure, you’d only have one button, but as Mac users will tell you, that’s not necessarily a problem.

The drawbacks to this method are that accurate eye-tracking hardware is expensive, cutting down on the spacebar’s size will be uncomfortable to many people (at least at first), and trying to control the mouse via eye-tracking is somewhat difficult, since eyes are prone to drifting and moving somewhat unsteadily. For more information on eye-tracking, see the Stanford HCI EyePoint project and the openEyes resources from Iowa State University. And finally, it leaves users with the suboptimal QWERTY keyboard layout, which most people won’t voluntarily abandon.

The big pro of this method is that it gives you the ability to handle positional input with your eyes as well as use both hands for typing on a full-size keyboard–a truly two-handed interface, as long as the tracking hardware and software can be made usably accurate.

But eye-tracking isn’t the only potential solution. Ever since Apple announced the iPhone, the web has been abuzz with talk of how multi-touch screens will revolutionize computers. Most touch-sensitive screens in the past have only been single-touch: they could only sense where you’re touching with one finger, and if you used more fingers, the interface would either ignore them or get confused. Multi-touch screens allow you to touch the screen with multiple fingers, manipulating objects from various points at the same time.

While people are oohing and aahing over Apple, most of them haven’t heard of an NYU researcher named Jefferson Han who has been working on an intuitive multi-touch interface for years. While his technology isn’t available outside his own lab yet, someday it will be. To get an idea of just how cool a well-designed multi-touch interface can be, try watching this video of Jeff demonstrating his interface at the TED Conference in 2006 or reading this in-depth FastCompany article (and be sure to watch the accompanying video for that, too).

Unfortunately, multi-touch screens don’t necessarily provide any sort of fantastic new method for text input. You can use both hands to manipulate objects on-screen, but when you want to type, you’ll be forced to take both hands off the screen and place them on a keyboard, or maybe just type on a simulated on-screen keyboard. Either way, you can’t effectively do both textual and kinetic input, so while multi-touch screens may provide a more intuitive and thorough way to interact with the computer kinetically, it still leaves us with the original problem.

I envision our redesigned computer using a combination of these two technologies. While typing, you can control the mouse with your eyes and a mouse-button on the keyboard. If you want to manipulate things on-screen more directly, you can reach up and do so via the multi-touch capability. This isn’t ideal, but I think it’s better than what we have now, and it’ll have to do until we get those direct neural interfaces up and running.

Next time, I’ll explore how the core of the computer’s operating system can be better-designed to support a human way of thinking.

(This post is the third part of my five-part post series “Reimagining the PC.” You might like to start back at the beginning.)

This entry was posted on Wednesday, Apr 18, 2007 at 7:40 am and is tagged design, technology, usability. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.