A web application for allowing local and remote control of automated
blinds.
Uses a set of four photographs of each room, with the blackout and screen shades all open or closed, to create a composite image showing the current state of the room, and allowing users to tap on the photo to specify the positions to which they want specific blinds to move. Also allows the definition of macros to quickly move groups of blinds to specific positions. Macros can be run globally or within individual rooms.
Provides support for guest access (control of specific rooms, or periodic activation for pet sitters).
In my setup, relies on Apache to terminate SSL using a LetsEncrypt-managed certificate, for a subdomain that is reverse proxied to a different port on localhost, so we only need to deal with ordinary HTTP communication.
Offers a web socket used by a daemon running on our home network which relays queries and commands to the blind controller software.
Incorporates astronomical algorithms and open weather data so that blinds can be closed on warm days during the part of the day where the sun is shining through them, and can allow you to go to sleep enjoying cityscape views, but close the blinds at astronomical dawn so you don't get awakened earlier than you want.
You will need Leiningen 2.0 or above installed, and a recent Java distribution, to build and run the Clojure web application.
The blind controller daemon requires python3 and the
websocket-client, edn_format, keyring, and
pyControl4 libraries.
Currently, it is unable to retry when the connection to the server is lost, so you need to wrap it in a mechanism that respawns it when it dies.
For weather support, you will need an API key from OpenWeather. To be notified when there are problems communicating with the blind controller daemon, you will need to set up an IFTTT web hook. The keys for both of these will go in your environment variables as described above.
To start a web server for the application, run:
lein run
For a deployed installation, you will want to build a standalone überjar, using:
lein uberjar
Then copy the resulting target/uberjar/shade.jar to somewhere like
/usr/local/lib/shade.jar, and configure it to be run as a system
daemon. Here is an example of a systemd configuration file which
does that on a modern Linux, shade.service:
# See https://github.com/brunchboy/shade
[Unit]
Description=The Shade Web Application
Wants=httpd-init.service
After=network.target remote-fs.target nss-lookup.target httpd-init.service
[Service]
Type=simple
Environment="DATABASE_URL=jdbc:postgresql:shade?user=shade&password=...elided..." "WEBSOCKET_TOKEN=...elided..." "OPENWEATHER_API_KEY=...elided..." "IFTTT_WEBHOOK_KEY=...elided..." "JAVA_TOOL_OPTIONS=-Xmx256m" "NREPL_PORT=7001"
SuccessExitStatus=143
ExecStart=/usr/bin/java -Dconf=/usr/local/etc/shade.edn -jar /usr/local/lib/shade.jar
KillMode=process
Restart=on-failure
User=james
Group=james
[Install]
WantedBy=multi-user.target
Because this file contains secrets such as passwords and tokens, you
should make sure it is readable only by root. And as you may have
noticed within it, there is a reference to a less-sensitive
configuration file that is read by the web server at startup. In my
case that is located at /usr/local/etc/shade.edn, and has the
following contents:
{:location {:latitude 43.07555555555556
:longitude -89.38611111111112
:timezone "America/Chicago"}
:cdn-url "https://d22vjwe5tlkwmh.cloudfront.net/shade"}This reflects the location and timezone to be used for astronomical
calculations, and the CloudFront content delivery network used to
efficiently serve the room images. You can skip using a CDN like that
and embed the images directly in the img subdirectory of the
resources folder, so they will be served by the web application
directly. If you do that, simply update the :cdn-url to point at
that path within your web application.
To start the blind controller daemon, which needs to run on the same
LAN as the Control4 Director appliance, you will need to set
environment variables C4_USERNAME and C4_DIRECTOR_IP to your
Control4 user name and the IP address of the Control4 Director
appliance on your local LAN. Your key ring will need to contain the
password for that account (under an entry for control4.com), and the
token you used passed as WEBSOCKET_TOKEN above as the password for
the account x-shade-token under an entry for
shade.deepsymmetry.org. Then you can launch the daemon using:
python3 src/python/client.py
Once again, you'll want to configure this to run automatically in the
background. Here is the macOS launchd agent file I use to run it as
an agent under my own login on a Mac Mini in our network rack, so it
has access to my system keychain for the above secrets,
~/Library/LaunchAgents/org.deepsymmetry.shades-agent.plist:
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
<plist version="1.0">
<dict>
<key>EnvironmentVariables</key>
<dict>
<key>C4_DIRECTOR_IP</key>
<string>192.168.1.150</string>
<key>C4_USERNAME</key>
<string>[email protected]</string>
<key>SHADE_WS_URL</key>
<string>wss://shade.my.domain</string>
</dict>
<key>KeepAlive</key>
<true/>
<key>Label</key>
<string>org.deepsymmetry.shaded</string>
<key>ProgramArguments</key>
<array>
<string>/usr/local/bin/python3</string>
<string>/Users/james/git/shade/src/python/client.py</string>
</array>
<key>RunAtLoad</key>
<true/>
<key>StandardErrorPath</key>
<string>/tmp/org.deepsymmetry.shaded.stderr</string>
<key>StandardOutPath</key>
<string>/tmp/org.deepsymmetry.shaded.stdout</string>
</dict>
</plist>Of course you will want to adjust the environment variable setting for
C4_DIRECTOR_IP, C4_USERNAME, and SHADE_WS_URL inside the
EnvironmentVariables dict to reflect your actual setup.
For development, you will want to bring up the application in a local
REPL for rapid testing of ideas. To facilitate that, you can put the
following content into the file dev-config.edn at the top level of
the project, and update it to reflect your setup. This is used only
for local development and is not committed to git.
As noted above, you can serve the room images directly from inside your web application rather than setting up a content distribution network to serve them if you want to keep things simple.
You will need to obtain your own free API key for OpenWeather, however. You can share that between development and production; the limits on a free account are far higher than Shade needs.
;; WARNING:
;;
;; The dev-config.edn file is used for local environment variables,
;; such as database credentials. This file is listed in .gitignore and
;; will be excluded from version control by Git.
{:dev true
:port 3000
;; when :nrepl-port is set the application starts the nREPL server on load
:nrepl-port 7000
:location {:latitude 43.07555555555556
:longitude -89.38611111111112
:timezone "America/Chicago"}
;; set your dev database connection URL here
:database-url "jdbc:postgresql://localhost/shade?user=shade&password=...elided..."
;; Set the URL from which large images and other resources are served here
:cdn-url "https://d22vjwe5tlkwmh.cloudfront.net/shade"
;; This controls the value of the special x-shade-token header that is required to open a web socket connection.
:websocket-token "...elided..."
;; The API key used to access OpenWeather forecast and current conditions information.
:openweather-api-key "...elided..."
}Once again, because this file contains secrets, you will want to make sure it is readable only by your account.
Released under the MIT license, http://opensource.org/licenses/MIT
Originally generated using Luminus version 4.38.
Copyright © 2022-2023 James Elliott