Oct
14

10/GUI and Reinventing More Than the Mouse

By Eric Shull  //  Blog Post  //  2 Comments

I stumbled onto 10/GUI over the weekend, a reimagining (not reinvention) of the traditional desktop. It has some interesting ideas, but most of all I’m glad to see someone is thinking about these things rather than just accepting the Old Ways.

10/GUI from C. Miller on Vimeo.

On Multitouch Keyboards

I find the keyboard in the video a bit awkward. Whereas we now use a keyboard and small trackpad or mouse, the multitouch input (which I shall refer to as the multipad) would swell our motion input device to the size of another keyboard. Personally, the keyboard-multipad combination looks like a keyboard from the ’80s, but that may just be me.

It would be great if the multipad could be used as the keyboard, rather than having a separate device. The question, then, is how do people type without tactile response. We’ve seen people type on multitouch keyboards before, but most of that is hunt and peck, not speed typing like keyboards allow. Having keys be graphical rather than tactile means the user has to look at the keyboard while they type, a huge slowdown for those of us who have the typing habit trained into muscle memory.

The good news is, typing is muscle memory for touch typists. I can type on a tabletop with no physical keyboard and hardly think about where my fingers are. The problem of course, is that I have no idea if I’m in the right place. On a real keyboard, the physical keys inform my fingers they have hit the spot. On the tabletop, I know where the keys are in a general way (which direction with which finger), but the exact location is elusive.

I suggest that if this technique is to be employed, that the location of the finger strike become unimportant, that the relationship between the finger’s original position and where it hit be taken into account instead. Thus, if my right pointer finger hits a position north of where it usually rests (on the ‘J’ key), chances are I meant to hit the ‘U’ key. Northwest from its usual position would be the ‘Y’ key. Who cares if my fingers are twice as long as average and I hit four inches north of ‘J’ (or my fingers are short and I hit a quarter inch north of ‘J’), the fact that I went north means I want the letter ‘U’.

I don’t remember the name of it now, but I used a keyboard for Windows Mobile that used a similar idea. It had only 9 keys. Tapping those keys would give you the most commonly used letters. The other letters were attained by starting on one of the 9 keys and moving in the appropriate direction. It took some getting used to, but ultimately I got to be fairly good with it. It was one of the programs I missed most when I switched to using an iPod Touch.

The Graphical Multitouch Keyboard

Now, what if instead of a mere multitouch input pad, there was some sort of graphical display incorporated into that? We’ve seen OLED keyboards already in which keys can change for different applications, but at the moment the technology is still somewhat expensive. I suggest an e-ink display, the same display used by Amazon’s Kindle e-reader. While still expensive, there’s no great need for a keyboard to be in color (look at what we’ve lived with to date) or to refresh its appearance many times a second. The huge advantage of using e-ink is the low power cost.

I’m just thinking out loud here.

  • Digg
  • del.icio.us
  • Facebook
  • Google Buzz
  • Google Reader
  • Reddit
  • NewsVine
  • RSS
  • Twitter
  • Technorati
  • StumbleUpon

2 Comments to “10/GUI and Reinventing More Than the Mouse”

  • I like your idea of having the keyboard work on a relative basis. I always wondered why those Bluetooth laser keyboards for PDAs/cell phones never worked like that.

    My biggest question regarding the 10/GUI video is: does having all of the windows in a horizontal array actually make the windows easier to find than our current random window model? They suggest it does, but I’m not sure I’m convinced. If you applied some GUI math for distance traveled I bet you could figure it out, but if the horizontal array wins, I bet it’s only by marginal means. I think accessing windows on the ends of the array would still be cumbersome.

Leave a comment