Re: 'QGMLWY' keyboard

Some time ago, someone used the genetic algorithm to come up with an
optimal layout and it ended up converging on dvorak. The original page
may be gone now, but it is referenced at this blog post:

http://keithdevens.com/weblog/archive/2002/Jul/06/EfficientKeyboardLayoutByGeneticAlgorithm

At CarpalX, their program did not converge onto dvorak. Why not? Did
it not take into account finger strength? Drumming motions? Or other
aspects? Or is CarpalX correct and it's Dvorak and the follow up
genetic algo experiment that have wrong or missing assumptions?

The following gentleman posts that he disagrees with Dvorak's
assumptions, but I'm skeptical as I do find alternating hands is
faster for me except for drumming motions like "t h". I'd also like to
see his 150wpm on video and measured by a program. There is also the
slight possibility that what's optimal for serious pianists when they
type is not optimal for the rest of us non-pianists. He uses thumbs on
letter keys. I don't think I could.

http://slashdot.org/comments.pl?sid=35481&cid=3832754


--
http://charles-esterbrook.com