Code related to the «Scrum vu des petites tranchées» talk to be given at Paris-Web 2012.

Clone the project:

$ git clone https://github.com/prswb/prswb.git
$ cd prswb

Create a new virtualenv in ./.env and enable it:

$ virtualenv --no-site-packages `pwd`/.env
$ source .env/bin/activate

Install dependencies:

$ pip install -r requirements-dev.txt

To launch a local dev webserver instance:

$ export UXPERIMENT_ENV=dev
$ python manage.py runserver
Validating models...

0 errors found
Django version 1.4.1, using settings 'uxperiment.settings'
Development server is running at http://127.0.0.1:8000/
Quit the server with CONTROL-C.

Head to http://127.0.0.1:8000/, profit.

This django project has a settings/ directory having a setting module per environment:

settings/
    |- base.py        # common shared settings
    |- dev.py         # dev settings
    |- production.py  # production settings (hosting platform yet to be determined)
    |- staging.py     # staging (hosted on heroku)
    |- test.py        # travis settings

Eventually, custom settings may be stored in a settings/local.py module.

The UXPERIMENT_ENV environment variable will set the specific settings module to load. To run the local webserver against a given environment:

$ UXPERIMENT_ENV=test python manage.py runserver

If you intend to work always with a given environment within the project virtualenv:

$ echo 'export UXPERIMENT_ENV=dev' >> .env/bin/postactivate
$ echo 'export unset UXPERIMENT_ENV' >> .env/bin/postdeactivate

The staging is hosted on Heroku and reachable at http://aqueous-mountain-3105.herokuapp.com/.

Note: You'll have to download and install the Heroku Toolbelt in order to manage some of the remote deployment procedures detailed below.

Your git user must have push privileges on the heroku repository. You must also provide your SSH public key to the admin account of the uxperiment project.

Once done, add this section to the .git/config of your local clone of the repo:

[remote "heroku"]
    url = git@heroku.com:aqueous-mountain-3105.git
    fetch = +refs/heads/*:refs/remotes/heroku/*

To push and deploy to heroku:

$ git push heroku master

After a push, Heroku will load the packages defined in the requirements.txt file. This file also contains packages for postgres and gunicorn.

At the end of the process, the Procfile will be used by Heroku to start the server.

To tell Heroku to use the staging settings module:

$ heroku config:set DJANGO_SETTINGS_MODULE=uxperiment.settings.staging

The staging settings module uses environment variables to configure the platform; to set them you have to use the heroku config:add command:

$ heroku config:add VARIABLE=VALUE

To run South migrations:

$ heroku run 'python manage.py migrate'

To tail the logs:

$ heroku logs --tail

The .travis.yml file at the root of the repository contains the required configuration.