Since keyboards have many more functions than the basic alphabet I prefer the current setup.

http://en.wikipedia.org/wiki/Chord_keyset

Actually, there is only 31 unique combinations.http://en.wikipedia.org/wiki/Combinatorial_mathematics#Combination_without_repetition

Which merely gets you 26 characters, plus shift, enter, space, delete, period.

The ASCII set is 95 printable characters, plus up to 33 non-printing characters.http://en.wikipedia.org/wiki/Ascii

To replace current keyboards, at least 7 keys will be necessary, but it is certainly a possibility. The test would be the gain in efficiency over the QWERTY model, but more training would be necessary because hunt and peck could not be used.

Good idea, but difficult to market.

So to replace

Sternographers and disabled folks use chording keyboards, and that's about the extent of the market. There have been many attempts over the years to sell chording keyboard to the general computer user, and all have been unsucessful to date.

Programmers and typists don't just use the keyboard for letter and numeric input. There's a wide array of data entry shortcuts and manipulations that can be performed on the keyboard to speed up both repetative data entry tasks and stylisitc formatting. Having MORE keys on the keyboard is a good thing, as long as they're arranged pragmatically (as in the DVORAK keyboard layout).

If you want to see the extent of the tools and shortcuts programmers use, go look at the vim or emacs text editors. I personally use vim, and the only equipment that could speed me up is having more keys to map command functions to.

I would love a 400-key, programmable keyboard.

Maybe the lack of visibility could be compensated somehow with the screen, which is something new. I'm thinking of some arrangement where the first keypress eliminates all impossible choices. If there are only 6 combinations employing "thumb", then as soon as "thumb" is pressed, all the others choices go away and only 6 thumb+? remain.

I'm sure there's a lot of work in this arena; anyone know what it's called?

Ron

Also, we're no longer limited to two positions per digit (up and down). There could easily be three horizontally, like rest, slid up, slid down.

Would that be faster? Maybe faster than a tiny QERTY with a stylus.

More compact? Well, possibly.

It might be fun to fiddle with -- but a tough, tough sell nonetheless.

Ron

Ok, one last comment: You might be interested in the work that's being done with fingerspelling for input.

Query: is there a fingerspelling for "F1" or "~" or "^"? Could there be? Probably have to have some fingerspelled equivalent of "Shift" or "Control" or "Command" that reinterprets the character that follows. Fun question, though.

Ron

You could kick out a lot more combinations through what arcade gamers call "sliding inputs". For example, hold down the first key, then quickly roll to the next key so you end up holding both keys for an instant before releasing the first key. You can also hold down a sequence of keys, and press the other keys in order to get certain effects. Using sequences of single presses, chords, slides, and chord-tapping.

A big problem with this method is that mistakes could be hard to control. Also, the learning curve would be extremely steep.

Very clever, those sliding inputs! Maybe audio feedback would cut the errors, like it does in gaming.

Of course, the OP was looking for "small." "Complete" and "fast" weren't on the list. Probably optimize as: "small, fast, and simple" for alphanumeric lowercase, then "small, fairly fast, fairly simple" for uppercase and common punctuation, then "small, slow, complex" for wierd punctuation, macros, etc.

You'd want to be able to enter alphanumeric lowercase entirely from memory, uppercase and common punctuation mostly from memory, with little 'cheat-blinks' at the screen to guide you, and the wierd stuff could be entirely screen-guided.