I've been thinking about morse code for the disabled lately. I know it sounds crazy but for any person with minimal movement...morse code is actually a good option. The code its self is setup so the shortest codes are the ones you're most likely to use. Morse is essentially a compressed code. Anyway, I was thinking about the paddles that my father used to use. It has two switches...hold down one and it spits out a steady stream of dits (dots), hold down the other and it spits out a steady stream of dahs (dashes). Then I kind of noticed that some of the timing aspects of morse...which make it easy to understand over static-filled radio are pointless on an input device for the disabled.

In morse code, dahs (dashes) and the gap between letters are 3 time units long. Dits (dots) and the gaps between the dits/dahs are 1 time unit long and finally the space between words is supposed to be around 7 time units long
Below we see a representation of the timing involved in the word "hello" with the normal between word spacing following.
X_X_X_X___X___X_XXX_X_X___XXX_XXX_XXX_______

...but on a computer interface, YOU know what button you pressed and so does the computer. The longer times for dahs are a waste. Also we could represent the between word gap (the space character) with a double key press. Below we can see "hello"'s timing characteristics under this system with O representing the double press...followed by the old one as a comparison
X_X_X_X___X___X_X_X_X___X_X_X_O X_X_X_X___X___X_XXX_X_X___XXX_XXX_XXX_______

On some of the longer codes, like say...sending "1000", the savings is even greater
X_X_X_X_X___X_X_X_X_X___X_X_X_X_X___X_X_X_X_X_O
X_XXX_XXX_XXX_XXX___XXX_XXX_XXX_XXX_XXX___XXX_XXX_XXX_XXX_XXX___XXX_XXX_XXX_XXX_XXX_______

The first example results in a 30% time savings, 53% in the second. But of course, I probably won't ever do anything with it...just goofing off, really. I suppose I should mention it to some non-profit group that deals with such things