build-your-own-radar: A CSS repository from davidmcgregor

A library that generates an interactive radar, inspired by thoughtworks.com/radar.

Demo

You can see this in action at https://radar.thoughtworks.com. If you plug in this data you'll see this visualization.

How To Use

The easiest way to use the app out of the box is to provide a public Google Sheet ID from which all the data will be fetched. You can enter that ID into the input field, sign in to Google using the prompt and your radar will be generated. The data must conform to the format below for the radar to be generated correctly.

Setting up your data

You need to make your data public in a form we can digest.

Create a Google Sheet. Give it at least the below column headers, and put in the content that you want:

name	ring	quadrant	isNew	description
Composer	adopt	tools	TRUE	Although the idea of dependency management ...
Canary builds	trial	techniques	FALSE	Many projects have external code dependencies ...
Apache Kylin	assess	platforms	TRUE	Apache Kylin is an open source analytics solution ...
JSF	hold	languages & frameworks	FALSE	We continue to see teams run into trouble using JSF ...

Sharing the sheet

In Google sheets, go to 'File', choose 'Publish to the web...' and then click 'Publish'.
Close the 'Publish to the web' dialog.
Copy the URL of your editable sheet from the browser (Don't worry, this does not share the editable version).

The URL will be similar to https://docs.google.com/spreadsheets/d/1waDG0_W3-yNiAaUfxcZhTKvl7AUCgXwQw8mdPjCz86U/edit. In theory we are only interested in the part between '/d/' and '/edit' but you can use the whole URL if you want.

Using CSV data

The other way to provide your data is using CSV document format. You can enter a publicly accessible URL (not behind any authentication) of a CSV file into the input field on the first page. For example, a raw URL for a CSV file hosted publicly on GitHub can be used. The format is just the same as that of the Google Sheet, the example is as follows:

name,ring,quadrant,isNew,description
Composer,adopt,tools,TRUE,"Although the idea of dependency management ..."
Canary builds,trial,techniques,FALSE,"Many projects have external code dependencies ..."
Apache Kylin,assess,platforms,TRUE,"Apache Kylin is an open source analytics solution ..."
JSF,hold,languages & frameworks,FALSE,"We continue to see teams run into trouble using JSF ..."

If you do not want to host the CSV file publicly, you can follow these steps to host the file locally on your BYOR docker instance itself.

Note: The CSV file parsing is using D3 library, so consult the D3 documentation for the data format details.

Using JSON data

Another other way to provide your data is using a JSON array. You can enter a publicly accessible URL (not behind any authentication) of a JSON file into the input field on the first page. For example, a raw URL for a JSON file hosted publicly on GitHub can be used. The format of the JSON is an array of objects with the the fields: name, ring, quadrant, isNew, and description.

An example:

[
  {
    "name": "Composer",
    "ring": "adopt",
    "quadrant": "tools",
    "isNew": "TRUE",
    "description": "Although the idea of dependency management ..."
  },
  {
    "name": "Canary builds",
    "ring": "trial",
    "quadrant": "techniques",
    "isNew": "FALSE",
    "description": "Many projects have external code dependencies ..."
  },
  {
    "name": "Apache Kylin",
    "ring": "assess",
    "quadrant": "platforms",
    "isNew": "TRUE",
    "description": "Apache Kylin is an open source analytics solution ..."
  },
  {
    "name": "JSF",
    "ring": "hold",
    "quadrant": "languages & frameworks",
    "isNew": "FALSE",
    "description": "We continue to see teams run into trouble using JSF ..."
  }
]

If you do not want to host the JSON file publicly, you can follow these steps to host the file locally on your BYOR docker instance itself.

Note: The JSON file parsing is using D3 library, so consult the D3 documentation for the data format details.

Building the radar

Paste the URL in the input field on the home page.

That's it!

Note: The quadrants of the radar, and the order of the rings inside the radar will be drawn in the order they appear in your data.

Check this page for step by step guidance.

More complex usage

To create the data representation, you can use the Google Sheet factory or CSV, or you can also insert all your data straight into the code.

The app uses Google Sheets APIs to fetch the data from a Google Sheet or D3.js if supplied as CSV, so refer to their documentation for more advanced interaction. The input data is sanitized by whitelisting HTML tags with sanitize-html.

The application uses webpack to package dependencies and minify all .js and .scss files.

By default, there is no distinction between both public and private Google Sheets as we now require authentication from Google (using OAuth Client ID and optionally, API Key), per Google's updated documentation. OAuth Client ID and API Key can be obtained from your Google developer console.

export CLIENT_ID=[Google Client ID]

Optionally, API Key can be set to bypass Google Authentication for public sheets.

export API_KEY=[Google API Key]

To enable Google Tag Manager, add the following environment variable.

export GTM_ID=[GTM ID]

Docker Image

We have released BYOR as a docker image for our users. The image is available in our DockerHub Repo. To pull and run the image, run the following commands.

$ docker pull wwwthoughtworks/build-your-own-radar
$ docker run --rm -p 8080:80 -e CLIENT_ID="[Google Client ID]" wwwthoughtworks/build-your-own-radar
$ open http://localhost:8080

Advanced option - Docker image with a CSV/JSON file from the host machine

You can check your setup by clicking on "Build my radar" and by loading the csv/json file from these locations:

$ docker pull wwwthoughtworks/build-your-own-radar
$ docker run --rm -p 8080:80 -e SERVER_NAMES="localhost 127.0.0.1" -v /mnt/radar/files/:/opt/build-your-own-radar/files wwwthoughtworks/build-your-own-radar
$ open http://localhost:8080

This will:

Spawn a server that will listen locally on port 8080.
Mount the host volume on /mnt/radar/files/ into the container on /opt/build-your-own-radar/files/.
Open http://localhost:8080 and for the URL enter: http://localhost:8080/files/${NAME_OF_YOUR_FILE}.${EXTENSION_OF_YOUR_FILE[csv/json]}. It needs to be a csv/json file.

You can now work locally on your machine, updating the csv/json file and render the result back on your browser. There is a sample csv and json file placed in spec/end_to_end_tests/resources/localfiles/ for reference.

Note:

If API Key is also available, same can be provided to the docker run command as -e API_KEY=[Google API Key].
For setting the publicPath in the webpack config while using this image, the path can be passed as an environment variable called ASSET_PATH.

Contribute

All tasks are defined in package.json.

Pull requests are welcome; please write tests whenever possible. Make sure you have nodejs installed.

git clone git@github.com:thoughtworks/build-your-own-radar.git
npm install
npm run quality - to run your tests
npm run dev - to run application in localhost:8080. This will watch the .js and .css files and rebuild on file changes

End to End Tests

To run End to End tests in headless mode

add a new environment variable 'TEST_URL' and set it to 'http://localhost:8080/'
add a new environment variable 'TEST_ENV' and set it to 'development' or 'production'
npm run test:e2e

To run End to End tests in debug mode

add a new environment variable 'TEST_URL' and set it to 'http://localhost:8080/'
add a new environment variable 'TEST_ENV' and set it to 'development' or 'production'
npm run start
Click on 'Run all specs' in cypress window

Note: Currently, end to end tests are not working, as the flow requires Google login via prompt, which Cypress does not support. We are working to find some alternative solution for this.

Don't want to install node? Run with one line docker

 $ docker run -p 8080:8080 -v $PWD:/app -w /app -it node:10.15.3 /bin/sh -c 'npm install && npm run dev'

Note: If you are facing Node-sass compile error while running, please prefix the command npm rebuild node-sass before npm run dev. like this

npm install && npm rebuild node-sass && npm run dev

After building it will start on localhost:8080

davidmcgregor/build-your-own-radar