Hmmm.... I've found some nice Marquardt keymodule switches[1], and
have gotten some samples. They are actuated using a linear rather
than curved movement like the snap switches that the other prototypes
use. I've used them to mock up a keyboard using corrugated cardboard
and hot-melt glue. Even though they have a 2.5oz actuation force
(double what John suggests), they feel quite nice.

I've come to two conclusions. First is that the "neutral" position of
the fingers changes depending on the position of the wrist. Relax
your hand completely and tip your wrist forward (apply no pressure to
the fingers). Your fingers will straighten out as the tendons wrap
farther around the corner your wrist has become. Now straighten your
wrist out (again, relax the fingers). Watch how your fingers curl
inwards. If you're having trouble doing this, move your wrist with
your other hand. *That* is the neutral position, with wrist straight.

All of the chordite prototypes I've seen so far show the fingers
straight. That seems to be wrong. In the neutral position, the
fingers are curved, and their planes of movement are not parallel.

Second conclusion is that every one of the keys needs to be pointed in
a different direction. The key actuated by the first phalange should
move in a plane whose normal points at the first knuckle. Then, since
the finger is naturally curved, the second phalange is moving in a
different plane than the first. So if the second key is to be in line
with that phalange's direction of movement, it must not be coplanar
with the first.

Thus, the only way to make an adjustable keyboard that actually
matches the neutral position of the hand is for every key to be
positionable with six degrees of freedom[2]. The keyboard prototypes
on the website are only adjustable in two degrees. Yes, six are
needed because the planes of movement of each finger intersect.
That's because the first knuckle allows for two degrees of freedom of
movement (you can spread your fingers out as in a baseball mitt.)

I think that's why John has had to fiddle so much with adjustable
keyboards -- because by making them adjustable he is constraining his
solution space to a sub-optimal position. If, instead, the keys go
underneath the phalanges' neutral positions (each in whatever
arbitrary position and angle is needed), I think a higher quality fit
will be achieved. Maybe even one which doesn't need to be made in
different sizes for different size hands!

[1] http://www.marquardt.de/produkte/baureihe.asp?bauid=88&sp=E
[2] Ahhhh, I love hot-melt glue and cardboard!

--
--my blog is at blog.russnelson.com | A computer without Python is
Crynwr sells support for free software | PGPok | like a CPU without memory:
521 Pleasant Valley Rd. | +1 315-323-1241 | it runs, but you can't do
Potsdam, NY 13676-3213 | Sheepdog | anything useful with it.