Resolves#70, Resolves#67
Still needs some regression testing in general, and a definite
regression is that rotary encoders are no longer (for the immediate time
being) supported.
Moves to a much simpler internal state tracking system, and FAR lighter
matrix scan.
Removes MicroPython support entirely.
What a short title for such a massive diff.
This (heavily squashed) commit adds support for Consumer keys such as
volume keys, media play/pause/stop, etc. by exposing four HID devices
over a single USB lane (as opposed to just exposing a keyboard). This
heavily refactors how HIDHelper works due to the new reporting
structure.
Many of the media keys were changed (mostly Keycodes.Media section), but
many (especially anything regarding Application keys) haven't been
touched yet - thus many keycodes may still be wrong. Probably worth
updating those soon, but I didn't get around to it yet. The definitive
list I refered to was
http://www.freebsddiary.org/APC/usb_hid_usages.php, which is basically
copy-pasta from the official USB HID spec at
https://www.usb.org/sites/default/files/documents/hut1_12v2.pdf
(warning: massive PDF, not light reading).
The only known regression this introduces is that instead of 6KRO as the
USB spec usually supports, we can now only have 5KRO (maybe even 4KRO),
for reasons I have yet to fully debug - this seems to be related to the
report having to include the device descriptor _and_ not supporting a
full 8 bytes as it used to. For now I'm willing to accept this, but it
definitely will be great to squash that bug.
This adds descriptor support for MOUSE and SYSCONTROL devices, as of yet
unimplemented.
This removes the need for the user to define... most things, honestly.
Notably, `main()` is no longer the end user's responsibility. This also
allows us to do fun stuff going forward like validating keymaps for
sanity (ex: the key assigned to `KC_MO(x)` should be assigned to
`KC_TRNS` on the target layer or the user will never be able to escape
that layer).
This also disambiguates `BOARD` to always refer to an actual slab of
silicon, renaming to `USER_KEYMAP`.
Entrypoints are now a bit more wild, and mostly-unsupported boards no
longer have working entrypoints. It's probably just time to scrap those
boards for now (until we have BLE HID and/or bitbang USB HID, at least).
