As a prerequisite to working on Pythondotorg, Docker, Docker Compose and make will need to be installed locally.


Docker Compose will be installed by Docker Mac and Docker for Windows automatically.

make is a build automation tool that automatically builds executebale programs and libraries from source code by reading files called Makefiles. The make utility comes defaulted with most unix distributions.

Getting started

To get the Pythondotorg source code, fork the repository on GitHub and clone it to your local machine:

git clone

Add a remote and sync regularly to stay current with the repository.

git remote add upstream
git checkout main
git fetch upstream 
git merge upstream/main 

Installing Docker

Install Docker Engine


The best experience for building Pythondotorg on Windows is to use the Windows Subsystem for Linux(WSL) in combination with both Docker for Windows and Docker for Linux.

Verify that the Docker installation is successful by running: docker -v

Running pythondotorg locally

Once you have Docker and Docker Compose installed, run:

make serve

This will pull down all the required docker containers, build the environment for pythondotorg, run migrations, load development fixtures, and start all of the necessary services.

Once complete, you will see the following in your terminal output:

web_1       | Starting development server at
web_1       | Quit the server with CONTROL-C.

You can view these results in your local web browser at: http://localhost:8000

To reset your local environment, run:

make clean

To apply migrations, run:

make migrate

To generate new migrations, run:

make migrations

You can also run arbitrary Django management commands via:

make manage <NAME_OF_COMMAND>

This is a simple wrapper around running python in the container, all arguments passed to make manage will be passed through.

Manual setup

First, install PostgreSQL on your machine and run it. pythondotorg currently uses Postgres 10.21.

Then clone the repository:

$ git clone git://

Then create a virtual environment:

$ python3.9 -m venv venv

And then you’ll need to install dependencies. You don’t need to use pip3 inside a Python 3 virtual environment:

$ pip install -r dev-requirements.txt

pythondotorg will look for a PostgreSQL database named pythondotorg by default. Run the following command to create a new database:

$ createdb pythondotorg -E utf-8 -l en_US.UTF-8


If the above command fails to create a database and you see an error message similar to:

createdb: database creation failed: ERROR:  permission denied to create database

Use the following command to create a database with postgres user as the owner:

$ sudo -u postgres createdb pythondotorg -E utf-8 -l en_US.UTF-8

Note that this solution may not work if you’ve installed PostgreSQL via Homebrew.

If you get an error like this:

createdb: database creation failed: ERROR:  new collation (en_US.UTF-8) is incompatible with the collation of the template database (en_GB.UTF-8)

Then you will have to change the value of the -l option to what your database was set up with initially.

To change database configuration, you can add the following setting to pydotorg/settings/ (or you can use the DATABASE_URL environment variable):

    'default': dj_database_url.parse('postgres:///your_database_name'),

If you prefer to use a simpler setup for your database you can use SQLite. Set the DATABASE_URL environment variable for the current terminal session:

$ export DATABASE_URL="sqlite:///pythondotorg.db"


If you prefer to set this variable in a more permanent way add the above line in your .bashrc file. Then it will be set for all terminal sessions in your system.

Whichever database type you chose, now it’s time to run migrations:

$ ./ migrate

To compile and compress static media, you will need compass and yui-compressor:

$ gem install bundler
$ bundle install


To install yui-compressor, use your OS’s package manager or download it directly then add the executable to your PATH.

To create initial data for the most used applications, run:

$ ./ create_initial_data

See create_initial_data for the command options to specify while creating initial data.

Finally, start the development server:

$ ./ runserver

Optional: Install Elasticsearch

The search feature in uses Elasticsearch engine. If you want to test out this feature, you will need to install Elasticsearch.

Once you have it installed, update the URL value of HAYSTACK_CONNECTIONS settings in pydotorg/settings/ to your local ElasticSearch server.

Generating CSS files automatically

Due to performance issues of django-pipeline, we are using a dummy compiler pydotorg.compilers.DummySASSCompiler in development mode. To generate CSS files, use sass itself in a separate terminal window:

$ cd static
$ sass --compass --scss -I $(dirname $(dirname $(gem which susy))) --trace --watch sass/style.scss:sass/style.css

Running tests

To run the test suite:

$ ./ test

To generate coverage report:

$ coverage run test
$ coverage report

Generate an HTML report with coverage html if you like.

Useful commands

  • Create a super user (for a new DB):

    $ ./ createsuperuser
  • Want to save some data from your DB before nuking it, and then load it back in?:

    $ ./ dumpdata --format=json --indent=4 $APPNAME > fixtures/$APPNAME.json