Development and Contributing
NOTE: This is a work in progress.
Everybody is invited and welcome to contribute to the smart mirror. There is a lot to do... If you are not a developer perhaps you would like to:
- Localize the smart mirror in a new language language (or improve an existing one)
If you are a developer and have a feature/capability you'd like to see, why not spend a couple of hours and help build it?
The process is straight-forward.
- create a branch from the dev branch following the following naming convention
initials/feature-namefor example if Evan Cohen was creating a feature he would name the branch
- Write the code for your feature/capability.
devbranch for features that are actively in development. If you would like to contribute please follow the contribution guidelines. To launch the mirror with a debug window attached use the following command:
npm start dev
The app directory contains the core of the smart-mirror:
js/core js of the smart mirror (bootstrapping angular, etc)
css/all of the smart mirror styles
locales/all of the localized speech commands (more on this later)
fonts/nobody knows what this folder does, but it's somehow necessary ;)
The plugins directory contains all of the plugins included in the smart mirror. A plugin consists of an optional combination of the following:
config.schema.jsondefines the configuration schema for your plugin. You can look at example schema and test your own at https://smart-mirror.io/playground/
index.htmlthe html partial for your plugin. This will be added to the main index.html as part of the new plugin location configuration.
controller.jsAll of the angular controller logic (data binding) for your plugin. This will be added to the main index.html as part of the new plugin location configuration.
service.jsshould your plugin need an angular service this will be added to the main index.html as part of the new plugin location configuration.
plugin.cssplugin specific classes for styles
locales\*.json- this folder contains the plugin specific translations required for both configuration )the config.schema.json file and plugin runtime (index.html) , * starting in V 0.28
The remote directory contains the code that powers the client configuration page(s). The "server" side code can be found in
Within this directory you will find the core
JSONlocalization files for the mirror. The mirror uses i18n to
$translatestrings rendered in the mirror (examples of this in
When making changes to these files make sure that you add string keys to all localization files, not just to the one that you speak.
Contains various help scripts. There aren't a ton today and they will expand over time.
The dev console is your friend! You can set breakpoints, log messages, and inspect the DOM - just like you would in a regular webdev console.
Since speech query limitations have becoming a limeting factor of development I've created a shim for Annyang to "simulate" speech to test the mirror. In the dev console try some of these examples:
// Play YouTube video
annyang.trigger("show me how to tie a bowtie");
// Play a song on SoundCloud
annyang.trigger("SoundCloud play Kero One so seductive");
// Display a map
annyang.trigger("show me a map of Seattle Washington");
I typically wouldn't recommend developing directly on the Pi (unless you are trying to debug a Pi specific issue, in which case, I'm sorry. It's not the fastest thing in the world.)
The mirror is compatible with linux and OSX, and developing there will be much nicer. You can plug your mirror into your computer and use it as a second (or third, you lucky duck) monitor. There's some logic in
main.jsthat will automatically put the electron app on to your secondary monitor.
It is not suggested that you develop on windows, the mirror is incompatible due to limitations of Snowboy. However, you can use a VM with ubuntu on it if you must...
However, if u want to use Windows as a development environment, you can you can use the WinScp or bitvise ssh clients to access the PI disk over SSH (make sure to enable it on the pi) and bot tools provide a file manager type view over the PI disk, so you can double click edit with your favorite Windows editor..
For Linux and Mac you can do the same with their file managers or addons. Like cyberduck for OSX or caja for linux