Useful information around the development process

Run Buster.JS

To run a manual test open a new terminal, change to your development directory of Buster.JS, execute the shell script (or batch file) to set the PATH and NODE_PATH, change to the directory of the example project and run the test, for example just by buster-test, depending on which tests and what kind of tests (browser or node) you want to run and how you want them to run (for example buster-static for browser tests).


Before you run Buster.JS, you should run buster-dev-tools clear and then buster-dev-tools deps. The first command deletes the node_modules folder of all projects (except for project buster-dev-tools). The second one installs the non-buster dependencies for all projects. That way you get back the state you have had right after setting up the development environment, concerning to the dependencies. Your changes to the source code are still there, of course.

This approach is especially important for the final manual test you make, just before you commit your changes.

Run tests

During the setup of the development environment via buster-dev-tools pull only the dependencies specified under dependencies in the package.json file are installed, not the devDependencies and not the optionalDependencies. They are not needed to run Buster.JS, but to run the tests. Therefore they have to be installed first. You can do this by changing to the root directory of the project you want to run the tests for and typing npm install.


The installation of the devDependencies can lead to problems, if a package has to be compiled, but you haven’t established a build toolchain for node-gyp. Read the installation chapter of node-gyp to learn more about this.

If you get the message Error: spawn ENOENT for the step node-gyp configure, try to set the paths for python as follows:


PYTHONPATH must point to your actual installation directory of Python 2.7.x, of course.

To run the tests type npm test in the root directory of the project.


You shouldn’t run the tests in an environment, where you have adjusted NODE_PATH and PATH for development as described here. There are two main reasons for that:

  1. Most of the projects use Buster.JS (or parts of it) for testing, which means, Buster.JS (or parts of it) will be installed in the node_modules directory of the project. Having NODE_PATH and PATH adjusted for development could lead to weird behaviour, because both versions of Buster.JS (or parts of it) are mixed up together.
  2. Running the test without adjusting NODE_PATH and PATH corresponds to how the tests run on Travis CI.


A comfortable way to debug Node applications is to use node-inspector. If you haven’t it installed yet, install it by:

npm install -g node-inspector

Open a terminal and start the inspector:


Debugging a Buster.JS run

Open the file buster/bin/buster-test and change the first line from #!/usr/bin/env node to #!/usr/bin/node --debug-brk.

Open another terminal, change to your development directory of Buster.JS, run the shell script (or batch file) to set the PATH and NODE_PATH, change to the example project and run the test by buster-test. The execution will be halted at the first instruction and you will get the info “debugger listening on port 5858”, the same as for debugging an automated test run. From that point on there is no difference between debugging an automated and a manual test run.


If you want to run other command than buster-test, for example buster-static, you have to edit the corresponding file instead.

Debugging a test run

To debug a test run just run npm run-script test-debug in the root directory of the project. The execution will be halted at the first instruction and you will get the info debugger listening on port 5858.

Open a Chrome browser, go to and you are ready to debug. More information about debugging with node-inspector can be found at node-inspector.

How to get your changes to the Buster.JS repositories

Because you don’t have write permission to the Buster.JS repositories you have to push your changes to a forked repository first and to create a pull request. One of the owner of the project will merge your changes into the corresponding Buster.JS repository at a later point, if we decide they are useful for the application.

This is one possible workflow:

  1. Create a fork of the repository you want to edit by clicking the “Fork” button on github.

  2. Open a terminal and change to the project you want to edit

  3. Type git remote -v and you should see something like this:

    origin (fetch)
    origin (push)

    This output is for the buster-cli repository and means, that we currently have only one remote directory connected to it and therefore can only fetch changes from there and can push our changes only to that repository, but unfortunately we don’t have write permission for it.

  4. Type git remote add fork<your_github_username>/<name_of_repository>.git


    If you open the forked repository on your Github account you will find the Url next to it. You can copy and paste it to prevent mistyping.

  5. Type git remote -v again and this time you should see something like this:

    fork<your_github_username>/buster-cli.git (fetch)
    fork<your_github_username>/buster-cli.git (push)
    origin (fetch)
    origin (push)
  6. Now you can specify to which remote repository you want to push your changes. Type git push fork master to push the changes to the master branch of the forked repository.

  7. Go to the forked repository on your Github account and press the “Compare and review” button to verify your changes and to create a pull request.

Finding your way around the code

We know that it’s hard to navigate the source code of Buster.JS if you are new at the project, because of the amount of projects/repositories. But hold on and don’t give up and you will see soon, that it’s not that complicated. A good point to start is to have a look at the Architecture overview, especially at the example.