Coding Conventions and Guidelines

Here you find an overview of our conventions on coding and version control.

Also have a look at our Contribution Guidelines.


  • We use GitHub Flow

  • Keep the main branch stable, as releases are derived from it. When contributing a breaking change, like API changes, inform the team about the intended merge.

  • When contributing code, create a Merge Request, and assign the repository maintainer for review. Do not squash commits when performing the merge.

  • We write changelogs in the keep a changelog format. Remind yourself to update the before committing code. Remember that changelogs are meant for humans and should reflect the end-user value. When doing a release we use the changelog as our Release Notes. When providing code via a Merge Request the maintainer will remind you to update the changelog too :-)

  • Avoid crunching commits and rebasing; set git config pull.rebase false to use recursive as your default merge strategy

  • We use conventional commits

  • Mention related ticket IDs where applicable, like #123 or play#123 when cross-referencing between repositories. For example you can close tickets by using closes #123.

  • Use atomic commits



We use Black with default settings, enforced by Flake8.

We use the default settings, except for a maximal line-length of 99 characters. If you are using a Black IDE integration, think about adapting its settings.

For code documentation we use the Flake8 Docstrings extension with Google-style formatting enabled (docstring-convention=google).

ECMAScript and TypeScript

We use ESLint as the common Linter. When your code is based on VueJS or Svelte, install the official IDE extensions. These extensions provide additional style checks and auto-formatting options.


We utilize an API-first approach. APIs are specified using OpenAPI 3. Find the API at

All the main aspects are documented within the spec. In some cases you may need some additional documentation in the docs. For example the API: Schedules and conflict resolution document can be found in “Developer Guide -> Misc”.


At the moment only Engine API is based on API-first. Steering API and Tank API are momentarily generated out of the code base.


When providing new features or refactoring code, please provide test cases. We do not expect 100% test coverage but are aiming for having the most important usage scenarios covered by automated CI/CD test suites.

Test and Demo Instances

We have two instances which are automatically and transparently deployed using GitLab:

  • - Developer/Testing instance holding the latest codebase of the main branches.

  • - Demo instance with a stable release, based on a release tag configured in GitLab.

We use the gitlab-ci for Deploying specific versions of aura.

  • You Can adjust the Container versions at gitlab-ci

       AURA_TANK_VERSION: "main"
  • you can choose between Docker Registry or Gitlab-Registry at gitlab-ci

    • DOCKER_REGISTRY: "docker" for Docker hub

    • DOCKER_REGISTRY: "gitlab"for Gitlab registry

After committing your desired changes, Gitlab auto deploy this version. Note that the Pipeline will only run on protected branches.

When a new Tagged Release is published, Gitlab will deploy this release.

There is a manual apply before deploy to ensure all Container Images with this Tag are published. Apply then Docker Images are released!

Auto Deployment based on feature branches

While this is currently not used, the Auto Deployment feature also provides custom builds based on feature branches.

For all components, feature branches automatically build Docker containers and push them to the gitlab-registry.

To roll out new features, the Compose-File, must be set to the corresponding tag of the feature branch.

After the commit, the version will be rolled out automatically.


The general documentation is located in aura/docs and hosted at When working on any component, also check if this documentation has to be updated or extended.

Developer Installation

For development, we recommend the native installation as outlined in the README of individual repositories.

Docker Compose Deployments

For production we highly recommend to run AURA using Docker and Docker Compose as outlined in the Administration Guide. But also as a developer you can benefit from the ease of an Docker Compose installation. For example when developing some Engine feature, you may want to run AURA Web with Docker Compose while testing.

Prepare your Development Environment

It is recommended to clone all projects for example in such folder structure:


Order of configuration

After that, you need to configure the projects in following order:

Web Projects

  1. Steering - Administration interface for schedules and programme information.

  2. Dashboard - Frontend to manage schedules, program info and audio files.

  3. Tank - Upload, pre-processing and storage of the audio files.

Play-out Projects

  1. Engine Core - Playout-engine to deliver the actual radio to the audience.

  2. Engine API - API Server to provide playlogs and information for the studio clock.

  3. Engine - Scheduling and remote control for the playout-engine.

Configuring the OpenID Clients

Dashboard and Tank authenticate against Steering using OpenID. We use OpenID Connect (OIDC) to implement OpenID.

Check out the OIDC configuration page, on how to get them talk to each other.

Component Documentation

For more detailed documentation read the README files in the individual repositories.

You get an overview of all repositories at