path: root/readme.md

# Quantum Mechanical Keyboard Firmware

[![Build Status](https://travis-ci.org/jackhumbert/qmk_firmware.svg?branch=master)](https://travis-ci.org/jackhumbert/qmk_firmware)

This is a keyboard firmware based on the [tmk_keyboard firmware](http://github.com/tmk/tmk_keyboard) with some useful features for Atmel AVR controllers, and more specifically, the [OLKB product line](http://olkb.com), the [ErgoDox EZ](http://www.ergodox-ez.com) keyboard, and the [Clueboard product line](http://clueboard.co/).

## Official website

For an easy-to-read version of this document and the repository, check out [http://qmk.fm](http://qmk.fm). Nicely formatted keyboard and keymap listings are also available there, along with the ability to download .hex files instead of having to setup a build environment and compile them.

## Included Keyboards

* [Planck](/keyboards/planck/)
* [Preonic](/keyboards/preonic/)
* [Atomic](/keyboards/atomic/)
* [ErgoDox EZ](/keyboards/ergodox_ez/)
* [Clueboard](/keyboards/clueboard/)
* [Cluepad](/keyboards/cluepad/)

The project also includes community support for [lots of other keyboards](/keyboards/).

## Maintainers

QMK is developed and maintained by Jack Humbert of OLKB with contributions from the community, and of course, [Hasu](https://github.com/tmk). This repo used to be a fork of [TMK](https://github.com/tmk/tmk_keyboard), and we are incredibly grateful for his founding contributions to the firmware. We've had to break the fork due to purely technical reasons - it simply became too different over time, and we've had to start refactoring some of the basic bits and pieces. We are huge fans of TMK and Hasu :)

This documentation is edited and maintained by Erez Zukerman of ErgoDox EZ. If you spot any typos or inaccuracies, please [open an issue](https://github.com/jackhumbert/qmk_firmware/issues/new).

The OLKB product firmwares are maintained by [Jack Humbert](https://github.com/jackhumbert), the Ergodox EZ by [Erez Zukerman](https://github.com/ezuk), and the Clueboard by [Zach White](https://github.com/skullydazed).

## Documentation roadmap

This is not a tiny project. While this is the main readme, there are many other files you might want to consult. Here are some points of interest:

* The readme for your own keyboard: This is found under `keyboards/<your keyboards's name>/`. So for the ErgoDox EZ, it's [here](keyboards/ergodox_ez/); for the Planck, it's [here](keyboards/planck/) and so on.
* The list of possible keycodes you can use in your keymap is actually spread out in a few different places:
  * [doc/keycode.txt](doc/keycode.txt) - an explanation of those same keycodes.
  * [quantum/keymap.h](quantum/keymap.h) - this is where the QMK-specific aliases are all set up. Things like the Hyper and Meh key, the Leader key, and all of the other QMK innovations. These are also explained and documented below, but `keymap.h` is where they're actually defined.
* The [TMK documentation](doc/TMK_readme.md). QMK is based on TMK, and this explains how it works internally.

# Getting started

Before you are able to compile, you'll need to install an environment for AVR development. You'll find the instructions for any OS below. If you find another/better way to set things up from scratch, please consider [making a pull request](https://github.com/jackhumbert/qmk_firmware/pulls) with your changes!

## Build Environment Setup

### Windows (Vista and later)
1. If you have ever installed WinAVR, uninstall it.
2. Install [MHV AVR Tools](https://infernoembedded.com/sites/default/files/project/MHV_AVR_Tools_20131101.exe). Disable smatch, but **be sure to leave the option to add the tools to the PATH checked**.
3. Install [MinGW](https://sourceforge.net/projects/mingw/files/Installer/mingw-get-setup.exe/download). During installation, uncheck the option to install a graphical user interface. **DO NOT change the default installation folder.** The scripts depend on the default location.
4. Clone this repository. [This link will download it as a zip file, which you'll need to extract.](https://github.com/jackhumbert/qmk_firmware/archive/master.zip) Open the extracted folder in Windows Explorer.
5. Double-click on the 1-setup-path-win batch script to run it. You'll need to accept a User Account Control prompt. Press the spacebar to dismiss the success message in the command prompt that pops up.
6. Right-click on the 2-setup-environment-win batch script, select "Run as administrator", and accept the User Account Control prompt. This part may take a couple of minutes, and you'll need to approve a driver installation, but once it finishes, your environment is complete!
7. Future build commands should be run from the MHV AVR Shell, which sets up an environment compatible with colorful build output. The standard Command Prompt will also work, but add `COLOR=false` to the end of all make commands when using it.

### Mac
If you're using [homebrew,](http://brew.sh/) you can use the following commands:

    brew tap osx-cross/avr
    brew install avr-libc
    brew install dfu-programmer

This is the recommended method. If you don't have homebrew, [install it!](http://brew.sh/) It's very much worth it for anyone who works in the command line.

You can also try these instructions:

1. Install Xcode from the App Store.
2. Install the Command Line Tools from `Xcode->Preferences->Downloads`.
3. Install [DFU-Programmer][dfu-prog].

### Linux
Install AVR GCC, AVR libc, and dfu-progammer with your favorite package manager.

Debian/Ubuntu example:

    sudo apt-get update
    sudo apt-get install gcc-avr avr-libc dfu-programmer

### Vagrant
If you have any problems building the firmware, you can try using a tool called Vagrant. It will set up a virtual computer with a known configuration that's ready-to-go for firmware building. OLKB does NOT host the files for this virtual computer. Details on how to set up Vagrant are in the [VAGRANT_GUIDE file](VAGRANT_GUIDE.md).

## Verify Your Installation
1. If you haven't already, obtain this repository ([https://github.com/jackhumbert/qmk_firmware](https://github.com/jackhumbert/qmk_firmware)). You can either download it as a zip file and extract it, or clone it using the command line tool git or the Github Desktop application.
2. Open up a terminal or command prompt and navigate to the `qmk_firmware` folder using the `cd` command. The command prompt will typically open to your home directory. If, for example, you cloned the repository to your Documents folder, then you would type `cd Documents/qmk_firmware`. If you extracted the file from a zip, then it may be named `qmk_firmware-master` instead.
3. To confirm that you're in the correct location, you can display the contents of your current folder using the `dir` command on Windows, or the `ls` command on Linux or Mac. You should see several files, including `readme.md` and a `quantum` folder. From here, you need to navigate to the appropriate folder under `keyboards/`. For example, if you're building for a Planck, run `cd keyboards/planck`.
4. Once you're in the correct keyboard-specific folder, run the `make` command. This should output a lot of information about the build process. More information about the `make` command can be found below.

# Customizing your keymap

In every keymap folder, the following files are recommended:

* `config.h` - the options to configure your keymap
* `keymap.c` - all of your keymap code, required
* `Makefile` - the features of QMK that are enabled, required to run `make` in your keymap folder
* `readme.md` - a description of your keymap, how others might use it, and explanations of features 

## The `make` command

The `make` command is how you compile the firmware into a .hex file, which can be loaded by a dfu programmer (like dfu-progammer via `make dfu`) or the [Teensy loader](https://www.pjrc.com/teensy/loader.html) (only used with Teensys). You can run `make` from the root (`/`), your keyboard folder (`/keyboards/<keyboard>/`), or your keymap folder (`/keyboards/<keyboard>/keymaps/<keymap>/`) if you have a `Makefile` there (see the example [here](/doc/keymap_makefile_example.mk)).

By default, this will generate a `<keyboard>_<keymap>.hex` file in whichever folder you run `make` from. These files are ignored by git, so don't worry about deleting them when committing/creating pull requests. Your .hex file will also be available on qmk.fm/keyboards/<keyboard>/keymaps/<keymap>/.

Below are some definitions that will be useful:

* The "root" (`/`) folder is the qmk_firmware folder, in which are `doc`, `keyboard`, `quantum`, etc.
* The "keyboard" folder is any keyboard project's folder, like `/keyboards/planck`.
* The "keymap" folder is any keymap's folder, like `/keyboards/planck/keymaps/default`.

Below is a list of the useful `make` commands in QMK:

* `make` - cleans automatically and builds your keyboard and keymap depending on which folder you're in. This defaults to the "default" layout (unless in a keymap folder), and Planck keyboard in the root folder
  * `make keyboard=<keyboard>` - specifies the keyboard (only to be used in root)
  * `make keymap=<keymap>` - specifies the keymap (only to be used in root and keyboard folder - not needed when in keymap folder)
* `make quick` - skips the clean step (cannot be used immediately after modifying config.h or Makefiles)
* `make dfu` - (requires dfu-programmer) builds and flashes the keymap to your keyboard once placed in reset/dfu mode (button or press `KC_RESET`). This does not work for Teensy-based keyboards like the ErgoDox EZ.
  * `keyboard=` and `keymap=` are compatible with this
* `make all-keyboards` - builds all keymaps for all keyboards and outputs status of each (use in root)
* `make all-keyboards-default` - builds all default keymaps for all keyboards and outputs status of each (use in root)
* `make all-keymaps [keyboard=<keyboard>]` - builds all of the keymaps for whatever keyboard folder you're in, or specified by `<keyboard>`
* `make all-keyboards-quick`, `make all-keyboards-default-quick` and `make all-keymaps-quick [keyboard=<keyboard>]` - like the normal "make-all-*" commands, but they skip the clean steps

Other, less useful functionality:

* `make COLOR=false` - turns off color output
* `make SILENT=true` - turns off output besides errors/warnings
* `make VERBOSE=true` - outputs all of the avr-gcc stuff (not interesting)

## The `Makefile`

There are 3 different `make` and `Makefile` locations:

* root (`/`)
* keyboard (`/keyboards/<keyboard>/`)
* keymap (`/keyboards/<keyboard>/keymaps/<keymap>/`)

The root contains the code used to automatically figure out which keymap or keymaps to compile based on your current directory and commandline arguments. It's considered stable, and shouldn't be modified. The keyboard one will contain the MCU set-up and default settings for your keyboard, and shouldn't be modified unless you are the producer of that keyboard. The keymap Makefile can be modified by users, and is optional. It is included automatically if it exists. You can see an example [here](/doc/keymap_makefile_example.mk) - the last few lines are the most important. The settings you set here will override any defaults set in the keyboard Makefile. **It is required if you want to run `make` in the keymap folder.**

### Makefile options

Set the variables to `no` to disable them, and `yes` to enable them.

`BOOTMAGIC_ENABLE`

This allows you to hold a key and the salt key (space by default) and have access to a various EEPROM settings that persist over power loss. It's advised you keep this disabled, as the settings are often changed by accident, and produce confusing results that makes it difficult to debug. It's one of the more common problems encountered in help sessions.

`MOUSEKEY_ENABLE`

This gives you control over cursor movements and clicks via keycodes/custom functions.

`EXTRAKEY_ENABLE`

This allows you to use the system and audio control key codes.

`CONSOLE_ENABLE`

TODO

`COMMAND_ENABLE`

TODO

`SLEEP_LED_ENABLE`

Enables your LED to breath while your computer is sleeping. Timer1 is being used here. This feature is largely unused and untested, and needs updating/abstracting.

`NKRO_ENABLE`

This allows for n-key rollover (default is 6) to be enabled. It is off by default, but can be forced by adding `#define FORCE_NKRO` to your config.h.

`BACKLIGHT_ENABLE`

This enables your backlight on Timer1 and ports B5, B6, or B7 (for now). You can specify your port by putting this in your `config.h`:

    #define BACKLIGHT_PIN B7

`MIDI_ENABLE`

This enables MIDI sending and receiving with your keyboard. To enter MIDI send mode, you can use the keycode `MI_ON`, and `MI_OFF` to turn it off. This is a largely untested feature, but more information can be found in the `quantum/quantum.c` file.

`UNICODE_ENABLE`

This allows you to send unicode symbols via `UC(<unicode>)` in your keymap. Only codes up to 0x7FFF are currently supported.

`BLUETOOTH_ENABLE`

This allows you to interface with a Bluefruit EZ-key to send keycodes wirelessly. It uses the D2 and D3 pins.

`AUDIO_ENABLE`

This allows you output audio on the C6 pin (needs abstracting). See the [audio section](#driving-a-speaker---audio-support) for more information.

### Customizing Makefile options on a per-keymap basis

If your keymap directory has a file called `Makefile` (note the filename), any Makefile options you set in that file will take precedence over other Makefile options for your particular keyboard.

So let's say your keyboard's makefile has `BACKLIGHT_ENABLE = yes` (or maybe doesn't even list the `BACKLIGHT_ENABLE` option, which would cause it to be off). You want your particular keymap to not have the debug console, so you make a file called `Makefile` and specify `BACKLIGHT_ENABLE = no`.

You can use the `doc/keymap_makefile_example.md` as a template/starting point.

## The `config.h` file

There are 2 `config.h` locations:

* keyboard (`/keyboards/<keyboard>/`)
* keymap (`/keyboards/<keyboard>/keymaps/<keymap>/`)

The keyboard `config.h` is included only if the keymap one doesn't exist. The format to use for your custom one [is here](/doc/keymap_config_h_example.h). If you want to override a setting from the parent `config.h` file, you need to do this:

```c
#undef MY_SETTING
#define MY_SETTING 4
```

For a value of `4` for this imaginary setting. So we `undef` it first, then `define` it.

You can then override any settings, rather than having to copy and paste the whole thing.

## Going beyond the keycodes

Aside from the [basic keycodes](doc/keycode.txt), your keymap can include shortcuts to common operations.

### Switching and toggling layers

`MO(layer)` - momentary switch to *layer*. As soon as you let go of the key, the layer is deactivated and you pop back out to the previous layer. When you apply this to a key, that same key must be set as `KC_TRNS` on the destination layer. Otherwise, you won't make it back to the original layer when you release the key (and you'll get a keycode sent). You can only switch to layers *above* your current layer. If you're on layer 0 and you use `MO(1)`, that will switch to layer 1 just fine. But if you include `MO(3)` on layer 5, that won't do anything for you -- because layer 3 is lower than layer 5 on the stack.

`OSL(layer)` - momentary switch to *layer*, as a one-shot operation. So if you have a key that's defined as `OSL(1)`, and you tap that key, then only the very next keystroke would come from layer 1. You would drop back to layer zero immediately after that one keystroke. That's handy if you have a layer full of custom shortcuts -- for example, a dedicated key for closing a window. So you tap your one-shot layer mod, then tap that magic 'close window' key, and keep typing like a boss. Layer 1 would remain active as long as you hold that key down, too (so you can use it like a momentary toggle-layer key with extra powers).

`LT(layer, kc)` - momentary switch to *layer* when held, and *kc* when tapped. Like `MO()`, this only works upwards in the layer stack (`layer` must be higher than the current layer).

`TG(layer)` - toggles a layer on or off. As with `MO()`, you should set this key as `KC_TRNS` in the destination layer so that tapping it again actually toggles back to the original layer. Only works upwards in the layer stack.


### Fun with modifier keys

* `LSFT(kc)` - applies left Shift to *kc* (keycode) - `S(kc)` is an alias
* `RSFT(kc)` - applies right Shift to *kc*
* `LCTL(kc)` - applies left Control to *kc*
* `RCTL(kc)` - applies right Control to *kc*
* `LALT(kc)` - applies left Alt to *kc*
* `RALT(kc)` - applies right Alt to *kc*
* `LGUI(kc)` - applies left GUI (command/win) to *kc*
* `RGUI(kc)` - applies right GUI (command/win) to *kc*
* `HYPR(kc)` - applies Hyper (all modifiers) to *kc*
* `MEH(kc)`  - applies Meh (all modifiers except Win/Cmd) to *kc*
* `LCAG(kc)` - applies CtrlAltGui to *kc*

You can also chain these, like this:

    LALT(LCTL(KC_DEL)) -- this makes a key that sends Alt, Control, and Delete in a single keypress.

The following shortcuts automatically add `LSFT()` to keycodes to get commonly used symbols. Their long names are also available and documented in `/quantum/keymap_common.h`.

    KC_TILD  ~
    KC_EXLM  !
    KC_AT    @
    KC_HASH  #
    KC_DLR   $
    KC_PERC  %
    KC_CIRC  ^
    KC_AMPR  &
    KC_ASTR  *
    KC_LPRN  (
    KC_RPRN  )
    KC_UNDS  _
    KC_PLUS  +
    KC_DQUO  "
    KC_LCBR  {
    KC_RCBR  }
    KC_LABK  <
    KC_RABK  >
    KC_PIPE  |
    KC_COLN  :

`OSM(mod)` - this is a "one shot" modifier. So let's say you have your left Shift key defined as `OSM(MOD_LSFT)`. Tap it, let go, and Shift is "on" -- but only for the next character you'll type. So to write "The", you don't need to hold down Shift -- you tap it, tap t, and move on with life. And if you hold down the left Shift key, it just works as a left Shift key, as you would expect (so you could type THE). There's also a magical, secret way to "lock" a modifier by tapping it multiple times. If you want to learn more about that, open an issue. :)

`MT(mod, kc)` - is *mod* (modifier key - MOD_LCTL, MOD_LSFT) when held, and *kc* when tapped. In other words, you can have a key that sends Esc (or the letter O or whatever) when you tap it, but works as a Control key or a Shift key when you hold it down.

These are the values you can use for the `mod` in `MT()` and `OSM()` (right-hand modifiers are not available for `MT()`):

  * MOD_LCTL
  * MOD_LSFT
  * MOD_LALT
  * MOD_LGUI
  * MOD_HYPR
  * MOD_MEH

These can also be combined like `MOD_LCTL | MOD_LSFT` e.g. `MT(MOD_LCTL | MOD_LSFT, KC_ESC)` which would activate Control and Shift when held, and send Escape when tapped.

We've added shortcuts to make common modifier/tap (mod-tap) mappings more compact:

  * `CTL_T(kc)` - is LCTL when held and *kc* when tapped
  * `SFT_T(kc)` - is LSFT when held and *kc* when tapped
  * `ALT_T(kc)` - is LALT when held and *kc* when tapped
  * `GUI_T(kc)` - is LGUI when held and *kc* when tapped
  * `ALL_T(kc)` - is Hyper (all mods) when held and *kc* when tapped. To read more about what you can do with a Hyper key, see [this blog post by Brett Terpstra](http://brettterpstra.com/2012/12/08/a-useful-caps-lock-key/)
  * `LCAG_T(kc)` - is CtrlAltGui when held and *kc* when tapped
  * `MEH_T(kc)` - is like Hyper, but not as cool -- does not include the Cmd/Win key, so just sends Alt+Ctrl+Shift.

### Space Cadet Shift: The future, built in

Steve Losh [described](http://stevelosh.com/blog/2012/10/a-modern-space-cadet/) the Space Cadet Shift quite well. Essentially, you hit the left Shift on its own, and you get an opening parenthesis; hit the right Shift on its own, and you get the closing one. When hit with other keys, the Shift key keeps working as it always does. Yes, it's as cool as it sounds.

To use it, use `KC_LSPO` (Left Shift, Parens Open) for your left Shift on your keymap, and `KC_RSPC` (Right Shift, Parens Close) for your right Shift. 

It's defaulted to work on US keyboards, but if your layout uses different keys for parenthesis, you can define those in your `config.h` like this:

    #define LSPO_KEY KC_9
    #define RSPC_KEY KC_0

You can also choose between different rollover behaviors of the shift keys by defining:

    #define DISABLE_SPACE_CADET_ROLLOVER

in your `config.h`. Disabling rollover allows you to use the opposite shift key to cancel the space cadet state in the event of an erroneous press instead of emitting a pair of parentheses when the keys are released.

The only other thing you're going to want to do is create a `Makefile` in your keymap directory and set the following:

```
COMMAND_ENABLE   = no  # Commands for debug and configuration
```

This is just to keep the keyboard from going into command mode when you hold both Shift keys at the same time.

### The Leader key: A new kind of modifier

If you've ever used Vim, you know what a Leader key is. If not, you're about to discover a wonderful concept. :) Instead of hitting Alt+Shift+W for example (holding down three keys at the same time), what if you could hit a _sequence_ of keys instead? So you'd hit our special modifier (the Leader key), followed by W and then C (just a rapid succession of keys), and something would happen.

That's what `KC_LEAD` does. Here's an example:

1. Pick a key on your keyboard you want to use as the Leader key. Assign it the keycode `KC_LEAD`. This key would be dedicated just for this -- it's a single action key, can't be used for anything else.
2. Include the line `#define LEADER_TIMEOUT 300` somewhere in your keymap.c file, probably near the top. The 300 there is 300ms -- that's how long you have for the sequence of keys following the leader. You can tweak this value for comfort, of course.
3. Within your `matrix_scan_user` function, do something like this:

```
void matrix_scan_user(void) {
  LEADER_DICTIONARY() {
    leading = false;
    leader_end();

    SEQ_ONE_KEY(KC_F) {
      register_code(KC_S);
      unregister_code(KC_S);
    }
    SEQ_TWO_KEYS(KC_A, KC_S) {
      register_code(KC_H);
      unregister_code(KC_H);
    }
    SEQ_THREE_KEYS(KC_A, KC_S, KC_D) {
      register_code(KC_LGUI);
      register_code(KC_S);
      unregister_code(KC_S);
      unregister_code(KC_LGUI);
    }
  }
}
```

As you can see, you have three function. you can use - `SEQ_ONE_KEY` for single-key sequences (Leader followed by just one key), and `SEQ_TWO_KEYS` and `SEQ_THREE_KEYS` for longer sequences. Each of these accepts one or more keycodes as arguments. This is an important point: You can use keycodes from **any layer on your keyboard**. That layer would need to be active for the leader macro to fire, obviously.

### Tap Dance: A single key can do 3, 5, or 100 different things

Hit the semicolon key once, send a semicolon. Hit it twice, rapidly -- send a colon. Hit it three times, and your keyboard's LEDs do a wild dance. That's just one example of what Tap Dance can do. It's one of the nicest community-contributed features in the firmware, conceived and created by [algernon](https://github.com/algernon) in [#451](https://github.com/jackhumbert/qmk_firmware/pull/451). Here's how Algernon describes the feature:

With this feature one can specify keys that behave differently, based on the amount of times they have been tapped, and when interrupted, they get handled before the interrupter.

To make it clear how this is different from `ACTION_FUNCTION_TAP`, lets explore a certain setup! We want one key to send `Space` on single tap, but `Enter` on double-tap.

With `ACTION_FUNCTION_TAP`, it is quite a rain-dance to set this up, and has the problem that when the sequence is interrupted, the interrupting key will be send first. Thus, `SPC a` will result in `a SPC` being sent, if they are typed within `TAPPING_TERM`. With the tap dance feature, that'll come out as `SPC a`, correctly.

The implementation hooks into two parts of the system, to achieve this: into `process_record_quantum()`, and the matrix scan. We need the latter to be able to time out a tap sequence even when a key is not being pressed, so `SPC` alone will time out and register after `TAPPING_TERM` time.

But lets start with how to use it, first!

First, you will need `TAP_DANCE_ENABLE=yes` in your `Makefile`, because the feature is disabled by default. This adds a little less than 1k to the firmware size. Next, you will want to define some tap-dance keys, which is easiest to do with the `TD()` macro, that - similar to `F()`, takes a number, which will later be used as an index into the `tap_dance_actions` array.

This array specifies what actions shall be taken when a tap-dance key is in action. Currently, there are two possible options:

* `ACTION_TAP_DANCE_DOUBLE(kc1, kc2)`: Sends the `kc1` keycode when tapped once, `kc2` otherwise.
* `ACTION_TAP_DANCE_FN(fn)`: Calls the specified function - defined in the user keymap - with the current state of the tap-dance action.

The first option is enough for a lot of cases, that just want dual roles. For example, `ACTION_TAP_DANCE(KC_SPC, KC_ENT)` will result in `Space` being sent on single-tap, `Enter` otherwise.

And that's the bulk of it!

Do note, however, that this implementation does have some consequences: keys do not register until either they reach the tapping ceiling, or they time out. This means that if you hold the key, nothing happens, no repeat, no nothing. It is possible to detect held state, and register an action then too, but that's not implemented yet. Keys also unregister immediately after being registered, so you can't even hold the second tap. This is intentional, to be consistent.

And now, on to the explanation of how it works!

The main entry point is `process_tap_dance()`, called from `process_record_quantum()`, which is run for every keypress, and our handler gets to run early. This function checks whether the key pressed is a tap-dance key. If it is not, and a tap-dance was in action, we handle that first, and enqueue the newly pressed key. If it is a tap-dance key, then we check if it is the same as the already active one (if there's one active, that is). If it is not, we fire off the old one first, then register the new one. If it was the same, we increment the counter and the timer.

This means that you have `TAPPING_TERM` time to tap the key again, you do not have to input all the taps within that timeframe. This allows for longer tap counts, with minimal impact on responsiveness.

Our next stop is `matrix_scan_tap_dance()`. This handles the timeout of tap-dance keys.

For the sake of flexibility, tap-dance actions can be either a pair of keycodes, or a user function. The latter allows one to handle higher tap counts, or do extra things, like blink the LEDs, fiddle with the backlighting, and so on. This is accomplished by using an union, and some clever macros.

In the end, let's see a full example!

```c
enum {
 CT_SE = 0,
 CT_CLN,
 CT_EGG
};

/* Have the above three on the keymap, TD(CT_SE), etc... */

void dance_cln (qk_tap_dance_state_t *state) {
  if (state->count == 1) {
    register_code (KC_RSFT);
    register_code (KC_SCLN);
    unregister_code (KC_SCLN);
    unregister_code (KC_RSFT);
  } else {
    register_code (KC_SCLN);
    unregister_code (KC_SCLN);
    reset_tap_dance (state);
  }
}

void dance_egg (qk_tap_dance_state_t *state) {
  if (state->count >= 100) {
    SEND_STRING ("Safety dance!");
    reset_tap_dance (state);
  }
}

const qk_tap_dance_action_t tap_dance_actions[] = {
  [CT_SE]  = ACTION_TAP_DANCE_DOUBLE (KC_SPC, KC_ENT)
 ,[CT_CLN] = ACTION_TAP_DANCE_FN (dance_cln)
 ,[CT_EGG] = ACTION_TAP_DANCE_FN (dance_egg)
};
```

### Temporarily setting the default layer

`DF(layer)` - sets default layer to *layer*. The default layer is the one at the "bottom" of the layer stack - the ultimate fallback layer. This currently does not persist over power loss. When you plug the keyboard back in, layer 0 will always be the default. It is theoretically possible to work around that, but that's not what `DF` does.

### Prevent stuck modifiers

Consider the following scenario:

1. Layer 0 has a key defined as Shift.
2. The same key is defined on layer 1 as the letter A.
3. User presses Shift.
4. User switches to layer 1 for whatever reason.
5. User releases Shift, or rather the letter A.
6. User switches back to layer 0.

Shift was actually never released and is still considered pressed.

If such situation bothers you add this to your `config.h`:

    #define PREVENT_STUCK_MODIFIERS

This option uses 5 bytes of memory per every 8 keys on the keyboard
rounded up (5 bits per key). For example on Planck (48 keys) it uses
(48/8)\*5 = 30 bytes.

### Remember: These are just aliases

These functions work the same way that their `ACTION_*` functions do - they're just quick aliases. To dig into all of the tmk ACTION_* functions, please see the [TMK documentation](https://github.com/jackhumbert/qmk_firmware/blob/master/doc/keymap.md#2-action).

Instead of using `FNx` when defining `ACTION_*` functions, you can use `F(x)` - the benefit here is being able to use more than 32 function actions (up to 4096), if you happen to need them.

## Macro shortcuts: Send a whole string when pressing just one key

Instead of using the `ACTION_MACRO` function, you can simply use `M(n)` to access macro *n* - *n* will get passed into the `action_get_macro` as the `id`, and you can use a switch statement to trigger it. This gets called on the keydown and keyup, so you'll need to use an if statement testing `record->event.pressed` (see keymap_default.c).

```c
const macro_t *action_get_macro(keyrecord_t *record, uint8_t id, uint8_t opt) // this is the function signature -- just copy/paste it into your keymap file as it is.
{
  switch(id) {
    case 0: // this would trigger when you hit a key mapped as M(0)
      if (record->event.pressed) {
        return MACRO( I(255), T(H), T(E), T(L), T(L), W(255), T(O), END  ); // this sends the string 'hello' when the macro executes
      }
      break;
  }
  return MACRO_NONE;
};
```
A macro can include the following commands:

* I() change interval of stroke in milliseconds.
* D() press key.
* U() release key.
* T() type key(press and release).
* W() wait (milliseconds).
* END end mark.

So above you can see the stroke interval changed to 255ms between each keystroke, then a bunch of keys being typed, waits a while, then the macro ends.

Note: Using macros to have your keyboard send passwords for you is possible, but a bad idea.

### Advanced macro functions

To get more control over the keys/actions your keyboard takes, the following functions are available to you in the `action_get_macro` function block:

* `record->event.pressed`

This is a boolean value that can be tested to see if the switch is being pressed or released. An example of this is

```c
if (record->event.pressed) {
  // on keydown
} else {
  // on keyup
}
```

* `register_code(<kc>);`

This sends the `<kc>` keydown event to the computer. Some examples would be `KC_ESC`, `KC_C`, `KC_4`, and even modifiers such as `KC_LSFT` and `KC_LGUI`.

* `unregister_code(<kc>);`

Parallel to `register_code` function, this sends the `<kc>` keyup event to the computer. If you don't use this, the key will be held down until it's sent.

* `layer_on(<n>);`

This will turn on the layer `<n>` - the higher layer number will always take priority. Make sure you have `KC_TRNS` for the key you're pressing on the layer you're switching to, or you'll get stick there unless you have another plan.

* `layer_off(<n>);`

This will turn off the layer `<n>`.

* `clear_keyboard();`

This will clear all mods and keys currently pressed.

* `clear_mods();`

This will clear all mods currently pressed.

* `clear_keyboard_but_mods();`

This will clear all keys besides the mods currently pressed.

* `update_tri_layer(layer_1, layer_2, layer_3);`

If the user attempts to activate layer 1 AND layer 2 at the same time (for example, by hitting their respective layer keys), layer 3 will be activated. Layers 1 and 2 will _also_ be activated, for the purposes of fallbacks (so a given key will fall back from 3 to 2, to 1 -- and only then to 0).

#### Naming your macros

If you have a bunch of macros you want to refer to from your keymap, while keeping the keymap easily readable, you can just name them like so:

```
#define AUD_OFF M(6)
#define AUD_ON M(7)
#define MUS_OFF M(8)
#define MUS_ON M(9)
#define VC_IN M(10)
#define VC_DE M(11)
#define PLOVER M(12)
#define EXT_PLV M(13)
```

As