* TKC1800: refactor
- tkc1800.h
- updated to use #pragma once
- visual alignment
- update config.h files to use #pragma once
- update all keymaps to use #include QMK_KEYBOARD_H
- delete redundant mod mask, KC_TRNS, and KC_NO definitions
- reformat keymap layers for readability
- add block comment for Function layer in default keymap
- delete commented action_get_macro code blocks (deprecated function)
- update files to better conform with QMK coding conventions
- update smt keymap to use set_single_persistent_default_layer (instead of persistent_default_layer_set)
- replace instances of `?=` with `=` in rules.mk
* TKC1800: Configurator visual fixes
- correct key sizes and positioning
- reformat file so each physical key object has its own line for readability
- update and add `label` keys for readability
* TKC1800: readme update
- fix image gallery link
- update Hardware Supported and Hardware Availability lines
- update make example (add :default suffix)
- update QMK Docs links (Newbs Guide; grammar)
* Add missing semicolons to smt keymap
This is the simplest, most efficient way I could come up with to silence
the "Failed to start write 60" error that occurs when QMK tries to talk
to a screen that doesn't exist.
iota_gfx_init passes a success boolean. We catch that into a global bool
(we could rewrite multiple functions to pass this as an argument, but
given the number of keyboards using this code it seemed less disruptive
this way) and then use that as a conditional on running the
iota_gfx_task in matrix_scan_user.
Tl;dr: if the screen doesn't init, the screen write code doesn't run.
