I’ve been playing with chorded input devices for years, and got the itch again recently.
I decided that I needed a decent chorder to put on my Spiffchorder electronics, and did a bit of shopping and design to that end. I bought a set of 10 Cherry MX Brown switches and “R2” shape caps from WASD Keyboards and some 0.060″ HDPE sheet from Sibe Automation via Amazon, which is exactly the right thickness to mount Cherry MX switches.
I went through 5 generations of layout prototypes CNC routed on my Shapeoko in EPS foam from a Heeks model until it was comfortable and interference free before investing real materials. That incurred a bit of delay while I sourced a single flute endmill (from my usual supplier to cut thin HDPE without melting it (see fail below). Mounting the switches is basically a matter of whacking out 14mm (yeah, mixed units, suck it) square holes with a 1.6mm endmill for a bit of corner radius. The lower lamination is currently just EPS foam milled (then knife cut to adjust for wires) to fit, eventually I’ll make a new one out of
cutting board inexpensively sourced 1/2″ thick HDPE stock with an internal cutout for the electronics, but have a board with a ATMega32u4 coming that I might rebase on first.
Some of the standard NASA/Alphadot (aka. BAT) chords are a bit mystifying, lots of 3-4-5 key chords for common letters, and w,y on single keys; I think there is a flow I haven’t quite mastered yet. It’s fucking hipster enough without messing with custom chordsets. You won’t be gaming (directionals suck even remapped) on one of these, and I suspect programming will always be awkward due to buried symbols, but I can see it for everyday text entry. The promise of two static hands on chorder and pointer (trackball) is as amazing as advertised (think The Mother of All Demos), especially when recumbent; as a related side-note, my raised, splayed hand gets cold easily, but it works just as well covered, as in “at my side, under the covers.”
This post was typed almost entirely on the pictured chorder as an
exercise in pain training experience. I’m slowly improving, but the claims of easy learning in the literature (60% of an individual’s qwerty speed in 5 hours on a precursor device) are overstated for the qwerty habituated.