###TODO
- Norm is really windows key
- Fn mode doesnt stick
- figure out actual keycodes and which is actual winkey
- add mouse support
- rust rewrite of the datahand Driver if I can't do the Haskell on the stm32 and see if I can just run rust on the teensy
- datahand combined with 443MHz radio + IR transmitter, for control of light, projector
- add a bunch of debug features by holding debug key on back + some letter to do various things like: debug+F enter flash mode, so i can avoid having to press through fabric debug+K: dump keymap as txt debug+L do a LED test / disco mode debug+R reset/reinit debug+D print debug info (the actual function of the debug button), listing line states debug[hold 5s] print out help as keystrokes, listing the above
-- TODO use some 2d rendering to make an image of the map, ideally one that looks similar to the original stickers. -- TODO does it make sense to leave ;: '" <> /? @^ #` -_ out of NAS? These keys are -- there because of original prog-dvorak placement. I get the utility of having -- their unshifted halves one keystroke away, but having some symbols accessed -- by shift+ and others by NAS+ may not be as conceptually clean as other -- solutions. -- Maybe add the shifter versions into side-keys in NAS like orig.datahand did? -- The orig layout added <> :; ,. /? (no |)which were all already accessible in Normal mode -- TODO DO 10k LAYER and support HEX ENTRY easily (have a-f, h as suffix, 0x prefix macro, ", 0x" macro maybe even -- TODO MACRO SUPPORT for situations like the above. -- TODO support exotic shift states to dhteensy like @^ key in prog.dv, maybe by adding a separate Shift key layer -- TODO maybe switch out @ for something used more frequently like #+* -- TODO BETTER FNMOUSE/ARROWS: have fn lever go to last fn state, then cycle fn -- types [arrows, mouse, more if desired]. Thereafter, use Norm to return, and -- then Fn to switch back to last Fn mode in that list. This frees up the two -- middle finger buttons for more useful binds. -- TODO Add FnArrow+NAS mode that turns arrow keys into pageup,down,home,end -- TODO naslock and capslock are pretty much wasted. i can hold the damn key -- down just fine, thanks. instead, make naslock go to 10kmode and capslock -- cycle whole layouts, maybe. i may want to put capslock somewhere buried in -- NAS, just in case i ever need it. -- TODO Profile switching: hold both locks and press a home row key to select -- that slot. or release a lock key last to cycle the opposite direction. -- TODO USB connection to other hand, attach as separate devices? usb3 micro connector has uart lines apparently? used in ergodox infinity this way, i read. -- TODO: NAS col, dqt, etc locations are same as Normal? -- * copy, paste, etc macros that are portable; alt-tab on single key on left-hand side -- * gnu screen: ctrl-a, ctrl-a n/p, " etc. -- * xmonad bindings -- TODO missing fn keys like printscreen,scrolllock, numlock -- TODO have winkey be both levers simulaneously pressed -- TODO C-a a,n,p,[,] macros for screen -- TODO give some thought to the difficulty i have typing ("1"); -- TODO FEATURE consider chording support for games. this is probably better/faster than hard to reach directional buttons. research what timing thresholds and behavior are most usable with chording. There is an inherent, unavoidable collision when mashing 2 buttons. Do I send both original keypresses anyway?(requires keymap accomodations to not send superfluous commands) or do i wait for a chord 50ms or so? (then all regular keys are unresponsive!) a middle ground would be to have some keys "chordable" meaning they would always exhibit the delay, and have others remain fully responsive. -- times when superflous keys would be okay: -- * modal menus in rtses, so long as underlying keys dont share binds -- * not many others, lol. this method seems pretty undesirable.