A cross platform library that forwards NetworkTables key/values over a WebSocket, so that you can easily write a Driver Station Dashboard for your robot in HTML5 + JavaScript.
This library does not provide a full dashboard solution, but is intended to provide the necessary plumbing for one to create one with only knowledge of HTML/JavaScript. Because the communications layer uses NetworkTables, you can connect to all FRC languages (C++, Java, LabVIEW, Python).
Note
NetworkTables is a protocol used for robot communication in the FIRST Robotics Competition, and can be used to talk to Shuffleboard/SmartDashboard. It does not have any security, and should never be used on untrusted networks.
Documentation can be found at http://pynetworktables2js.readthedocs.org/
- Download the latest pynetworktables2js.exe from GitHub at https://github.com/robotpy/pynetworktables2js/releases .
- Extract the exe from the zipfile, and copy it to your directory of HTML/JS files.
- Double click the exe to run it!
Note
By default, it will connect to 127.0.0.1. To connect to a robot,
you will need to pass the exe arguments to tell it where the robot is.
Use --help to see the available options.
Make sure to install python 3 on your computer, and on Windows you can execute:
py -3 -m pip install pynetworktables2js
On Linux/OSX you can execute:
pip install pynetworktables2js
Note
Technically, there's nothing stopping you from installing this on your robot, as there is a python interpreter available on the roboRIO (RobotPy). However, due to FRC bandwidth limitations, it's probably best to run the UI + server on your driver station laptop.
TL;DR: It's simpler.
pynetworktables2js lowers the barrier of entry for teams that want an additional way to tune/control their robot with a minimal amount of programming.
Lots of students and mentors know how to create simple web pages to display content, and there's lots of resources out there for creating dynamic content for webpages that use javascript. There is a lot of visually appealing content that others have created using web technologies -- why not leverage those resources to make something cool to control your robot?
You can just distribute your HTML files, and run a pynetworktables server using the following command from inside the directory:
python3 -m pynetworktables2js
Or on Windows:
py -3 -m pynetworktables2js
This will start a pynetworktables2js server using Tornado (which is installed by default) and it will serve the current directory. You can navigate your browser (I recommend Chrome) to http://127.0.0.1:8888 and see your website.
You will want to also pass either the --robot or --team switch:
py -3 -m pynetworktables2js --robot roborio-XXXX-frc.local py -3 -m pynetworktables2js --team XXXX
Dashboard mode currently doesn't work, as the underlying support in pynetworktables hasn't been implemented yet for the newer FRC Driver Station.
There are two example servers distributed with pynetworktables2js, one that uses tornado, and one that uses aiohttp. Either one should work.
Go to the 'example' directory distributed with pynetworktables2js, and run:
python3 tornado_server.py --robot 127.0.0.1
If you want to try this out with your current robot, you can do:
python3 tornado_server.py --robot roborio-XXX.local
If you are running pynetworktables2js on your driver station laptop, you can receive robot IP information directly from the Driver Station (handy during actual competitions):
python3 tornado_server.py --dashboard
If you navigate your browser (I recommend Chrome) to http://127.0.0.1:8888, all of the current NetworkTables values will be shown as they change.
One way of testing this out is use FIRST's TableViewer application (you can launch it using the "Outline Viewer" WPILib menu item in Eclipse), and start it in server mode.
Feel free to copy the example directory to create your own customized dashboard. Just add your custom files to the www directory.
pynetworktables2js is intended to be a project that all members of the FIRST community can quickly and easily contribute to. If you find a bug, or have an idea that you think others can use:
- Fork this git repository to your GitHub account
- Create your feature branch (
git checkout -b my-new-feature) - Commit your changes (
git commit -am 'Add some feature') - Push to the branch (
git push -u origin my-new-feature) - Create new Pull Request on GitHub
One place in particular I would love to see contributions is in adding useful JavaScript functions/objects that make creating dashboards even easier!
Leon Tan of FRC Team 1418 did the initial research/work to get this working, and created an initial working prototype for Team 1418's 2015 Dashboard, which was instrumental to winning an Innovation In Control award at the 2015 Greater DC Regional.
Dustin Spicuzza cleaned stuff up, rewrote things, added more functionality, wrote documentation, and packaged it so other teams could use it.