So... what's the maximum WPM one can achieve with this 5-finger piano keyboard thing? Of course, to be fair we'd probably have to double it if we were to compare it to that of a standard keyboard since it uses only one hand.
Various companies have tried to make five-finger chording keyboards over the years, but I have never seen one in person.
I strongly suspect that they don't really work very well. Hacking one together wouldn't be that hard, and yet you never see anyone but the diehard wearable-computing hobbyists using them. (The wearables folks, of course, really need the one-handed-typing feature.)
If they bought you so much as a 10% improvement you'd expect to find hobbyists using them and raving about them... just look at the adoption of Dvorak, or ask me about my awesome Kinesis keyboard.
The violin metaphor is a telling one: Despite their cost and the difficulty of learning to use them, violins and violinists aren't that hard to find. People take the trouble to learn violin because it pays off. So if you invent the chorded keyboard and twenty years later nobody has adopted it, you have to ask yourself whether you've really invented the equivalent of the violin. Perhaps you've invented the theremin instead: An instrument that is nigh-impossible to learn and that has no real repertoire beyond Good Vibrations and late-night monster movie soundtracks.
Stenographers use chorded keyboards, but the demand for speed is much higher. I think the primary concern is that typing speed simply isn't the bottleneck for most users. The learning curve for a regular keyboard might level off earlier, but few really need typing speeds in excess of what's attainable on a keyboard.
What's surprising, though, is that nobody sells a chorded keyboard for cellphones.
I think Emacs/vim are better examples of the violin. It's atrocious to start learning them, but once you do they're so much more powerful and efficient than a regular text editor.
You may still be able to find it secondhand. It has more than five buttons (12?), but they are chorded like Engelbart's design to access the 100-some necessary scan codes. Some users reported typing speeds approaching QWERTY, and supposedly with less RSI problems.
Edit: Wikipedia has links to a number of other prototypes and (mostly defunct) production models at the bottom of its chorded keyboard article:
The problem with RSI is that it's extremely personal. One person's wonder-working invention is another's torture device.
I don't doubt that there are many users who could benefit from a one-handed chorded keyboard -- I once met a programmer with the use of only one hand -- but many others will see no benefit, and others might have bigger RSI problems with chording than with regular keyboards.
Meanwhile, I find that the phrase "speeds approaching QWERTY" is not a very effective sales pitch. :)
I agree entirely (hence the various qualifiers in my post). The amazing WPM cited on the Twiddler page comes from a quote from Thad Starner, who is a prominent wearable computing researcher and one of the Twiddler's leading advocates. Some of his research on actual typing speeds, here:
> The problem with RSI is that it's extremely personal.
As someone who manages their RSI problems, I agree. My only bit of advice on this is if you find that you can pound out code for an extended bit of time and start to feel pain/soreness and after a little while recover...STOP - that is a major warning flag that I ignored tens of times - there is a cumulative effect/straw that breaks the camel's back, such that recovery time doesn't get you back to 100% you never really recover (baseline level of aching). That is RSI
http://web2.sys-con.com/read/536976_p.htm