Public | Automated Build

Last pushed: a year ago
Short Description
Dev image for GDQ Donation tracker. Experimental, so don't use this yet.
Full Description


Contains the settings and configuration for creating a simple deploy of the tracker.

In order to deploy the tracker, some boilerplate code is neccessary for configuration and management. The goal of this repository is to make doing so as simple as possible for any given user to get started developing on the tracker.

Getting a Working Copy of the Tracker

  1. Install Git. I'm assuming if you're here, you know enough about git and version control to get started. You can check if you have git with the command which git, and which version you have with git --version.
  2. Install Python. Currently, the tracker only supports version 2.6+ for now. Thus, getting whatever the latest 2.7 version is ideal. You can determine if Python is installed with the command which python, and which version of Python you have with the command python -V. Python version 3.x is not supported by our code at this time. If someone wants that to change, they will have to do the legwork themselves, we are not in a position to support both versions at this time.
  3. Install pip This is the package management system we use with the tracker, and its generally the best option for getting Python packages.
  4. Install node. Optional If you want to use or develop on the fancy new Javascript UI (which right now only supports the schedule editor) you'll need this. Right now only v5.x is supported, but others may work.
  5. Install direnv. Optional, Linux/OSX only This will help set up an isolated development environment.
  6. Clone this repository, typically I put it in a folder called donations, which is the path to which this repo will be referred for the remainder of these instructions:
    > git clone donations
  7. Make an empty directory under donations called db. This is where you can keep the working copy of your sqlite database. By default, the settings are such that one called db/testdb will be created, but of course you can modify that to suit your needs.
  8. Make a copy of under donations, and call it This is where you will enter any deployment-specific settings for your instance of the website.
    > cp
    1. (optional) Change the NAME field under the DATABASES variable to point at a different location if you wish.
    2. There are some other config variables related to timezone, e-mail, google docs, and giantbomb's API. None of these are neccessary to get started, and mostly can be ignored unless you are interested in that specific feature. Documentation on these fields is lacking, but it shouldn't be too diffficult to figure out how they work if you take a look at
  9. Clone the submodules.
    > git submodule update --init
    1. This will clone tracker (the main backend and the classic frontend), tracker_ui (the fancy new experimental Javascript UI), and GDQ's custom branch of django-paypal.
  10. Download the requirements in using pip:
    > pip install -r tracker/requirements.txt
    > pip install -r tracker_ui/requirements.txt
    1. If you are using Windows, you may need to delete the lines containing psycopg2 and chromium-compact-language-detector from tracker/requirements.txt (both are optional, and require compilation of C code, which is typically a hassle for people in a Windows environment).
    2. If you are under 'nix or Mac, you'll probably need to sudo this/run as administrator, unless you're using direnv as suggested above.
    3. The default pip configuration performs a fresh install of all packages, meaning that previously installed packages will be uninstalled before being reinstalled. If you get any exceptions during uninstallation, you can try the flag --ignore-installed to leave those packages alone and continue with other packages.
  11. Install the custom branch of django-paypal. Depending on how you have your Python paths set up, this may not be required.
    > pip install ./django-paypal
  12. Initialize the database. This app, and all the apps it depends on, can be initialized into the database using django's migrate command, python migrate. Notes:
    1. This is the actual command that will create the db/testdb file on your machine (if it does not exist already).
    2. If you are using a different location, or a different database type, you will need to make sure the permissions and settings are set up correctly.
    3. This is the general command to migrate all changes in the app. If you ever update any of the dependent libraries, or the tracker itself, you should run this command again.
  13. Create a superuser account for the admin with the command python createsuperuser and follow the prompts. This is the account you'll use to access your testing instance of the app.
  14. Install the required npm packages. Optional
    > cd tracker_ui && npm i
    Only needed if you want to use or develop the fancy new experimental UI.

Running the test server (see below on how to launch the UI)

You can run the test server, using the command python runserver [port]. The port argument is optional; the default is 8000.

You can navigate to the tracker at (where 8000 is the the port specified). To view the admin site, go to: and log in using the username/password you set up with createsuperuser.

Building the UI package (release mode)

Simply run the build command in the tracker_ui directory:
> webpack --config tracker_admin.release.webpack.js

This does two things:

  1. Builds the UI Javascript and CSS bundles and puts them in tracker_ui/static/gen.
  2. Outputs a manifest file to ui_admin.manifest.json so that Django knows where to find the resulting bundles.

This will allow the tracker UI to function, though if you want to develop with it you'll want to use the development proxy below, otherwise you'll not only have a minified build (difficult to debug!) but you'll have to rerun the command every time you make a change.

Running the UI development server

Webpack has a development server that can proxy requests to the backend. Once you've installed the required packages, you can run the server with the following command while in the tracker_ui folder:
> ./

It defaults to port 8080, so simply visit and you should be able to view the site just like the Django development server.

Note that if you change the port that the server is running on you'll need to edit shared.webpack.js to point to the correct port in the proxy section.

Server deployment

There are far too many different ways to deploy the server to go over every possibility here, so you should start with Deploying Django.

Note that node is NOT required to run the server in a production environment, it is ONLY needed to build the Javascript UI package. You can build it locally and simply copy the necessary files to your server. Don't forget the manifest file!

Docker (experimental, development environments only)

Alternately, you can use Docker. The packaged Dockerfile should build an isolated development environment for you, regardless of what OS you're on. While this is potentially the easiest to set up, it also has the most overhead. Note that this container is NOT intended for production use.

You can also download the latest image from Dockerhub.


If you have any questions, or would like some help getting set-up, or are interested in contributing, please don't hesitate to contact me at

Docker Pull Command

Comments (0)