One of my more popular open source projects is a Python weather forecasting package called Python-Forecastio.

The project wraps the Forecast.io JSON API in a Python interface. This provides an intuitive way to interact with the weather service.

Recently I have started testing the package on Travis. This is the first project which I have used Travis for testing, normally using CircleCI where open source transparency is not a concern.

My first impressions are quite good. You define your build configuration in a YAML file which should be placed in the root of your repo.

Here is mine.

language: python

python:
  - "3.3"
  - "2.7"

install:
  - "pip install -r requirements-test.txt"

script: nosetests --with-coverage

after_success: coveralls

The file specifies that the tests should be run against both Python 2.7 and 3.3. Testing both versions of Python also ensures cross compatibility.

The next step is to push a commit to your GitHub repo. Travis picks up the commit and runs the script command specified in the YAML file.

This can be a fairly slow process, it took around 15mins after pushing before I actually had the results. Especially when I’m used to the very quick turn around time of CircleCI.

Still, if your project is public, Travis will run your tests for free and this is a really good thing. Writing tests should be encoraged by the open source community because it sets a good example for novice developers and provides assurance to its user base. Travis is supporting this idea by making testing services more accessible.

If you want to take a look at my projects latest test results, they can be found here.