Public | Automated Build

Last pushed: 2 years ago
Short Description
Automated build for Tangerine-server. Why can't you make a current docker repository automated?
Full Description


This is one of two main Tangerine software repositories, the other is Tangerine-client. You will want to run this if you are planning on running a Tangerine server for editing assessments and uploading data from tablets.


The easiest way to get started is to create an Ubuntu 14.04 Droplet on Digital Ocean, SSH into it from a terminal, install Docker, and then run the following commands. You'll need the URL of the most recent release. Find that on the releases page here.

wget <latest release, choose tar.gz option>
tar xvf <the tarball>
cd <the now uncompressed tangerine folder>
# Edit the file to match your desired settings. Try `nano`. 
# You must edit all of the variables in the "Required to change" section. 

Now visit your Tangerine-server installation at the IP address or hostname of your installation.

If your server restarts or the container stops, you can later run the ./ script in the Tangerine-server folder. To upgrade your server, run the ./ script.

Develop on Tangerine

To develop on Mac or Windows, this project requires a working knowledge of docker and docker-machine. While you can issue Docker commands from Windows or Mac, Docker containers cannot run directly run on those platforms (yet) so it requires connecting to another machine running Linux using the docker-machine command. To learn Docker, check out the self-paced training on

Now that your docker command is connected to Linux machine it's time to get the code, but before you download the code, you will want to "fork it". Go to and click on the "Fork" button. Choose to fork the repository into your own personal Github account. From your forked version of the code, copy the clone URL and issue the following commands.

git clone <your clone url>
cd Tangerine-server

Now it's time to build and run your code. This first build will take up to 30 minutes depending on your Internet connection, processor power, and memory of your host machine. Future builds will take between 5 and 15 seconds because of the "Docker cache" you will be building on this first run.

# Edit to your desired setting. If you are using Docker Machine, make sure to set the T_HOST_NAME to the IP address of your Virtual Machine.

Now that you've built and run an image tagged as local, view your server from a web browser to confirm it is working. When you visit the site in a web browser, you will see output to your terminal indicating there is activity on the server.

You are now ready to start modifying code. We use the Edit-Build-Run workflow. That means everytime you edit code, you will build and run the code with the script. If you make an edit that breaks the build, you will see that in the output in your terminal. If all is well, you can stop following the logs with ctrl-c and it's now time to make a code change and see that reflected in your browser. For example, edit ./editor/app/_attachments/index.html and change the text in the <title> tag to be something like <title>Hello Tangerine</title>. If you reload your your browser you'll notice that the title tag has not changed. That's because the code your browser is viewing is based on the built image we made before the code change. To see our change run ./ again.

Now follow the logs again and check your browser. If all is well, you should now see the title of the web page as "Hello Tangerine".

Testing code inside of a container

This goes against the grain of how Docker ethos: you should not modify the code inside your docker instance. In this case,
we're making a trade-off due to the difficulties of developing this docker container locally.

Follow the instructions in the Developers section. Get into a running container to play around.

docker exec -it tangerine-server-container /bin/bash 

Go to the code for editor:

cd /tangerine-server/editor
npm start default

This npm command will run the default gulp command and watch for changes in your code.

SSH into another console to your server , docker exec into the same instance, edit your code.

Please note: you must create a new group when you wish to view your changes.

Be sure to commit your code ASAP. Once your container is gone; any uncommitted changes will also be gone.

Docker Pull Command