Skip to content

Contributing

Contributing to DUMB#

Thanks for helping improve Distributed Unlimited Media Bridge (DUMB).

Code, docs, testing feedback, and issue triage are all valuable contributions.

Branch Strategy#

  • dev is the default collaboration branch.
  • master is the production and release branch.
  • Open normal feature and fix PRs against dev.
  • Open PRs against master only for release work or approved hotfixes.

Development Setup#

Recommended: use the DUMB devcontainer:

Quick start:

  1. Fork I-am-PUID-0/DUMB.
  2. Clone your fork.
  3. Open the workspace in VS Code Dev Containers.
  4. Create a feature branch from dev.
  5. Make changes and run relevant checks.
  6. Open a PR to dev.

If you do not use Dev Containers, run with Docker using the same mounts and environment layout.

Pull Request Guidelines#

  • Keep PRs focused and small when possible.
  • Include a clear summary, motivation, and test notes.
  • Link related issues.
  • Add screenshots for UI changes.
  • Update docs when behavior changes.

Commit Format#

Use Conventional Commits for PR titles and commits, for example:

  • feat(api): add new symlink repair mode
  • fix(proxy): route iframe websocket path correctly
  • docs: update onboarding preflight guide

conventional-commits.yml validates PRs targeting both dev and master.

Dependabot and Automation Notes#

  • Dependabot updates are targeted to dev.
  • Integration testing for combined Dependabot updates runs from dev.
  • Release automation remains pinned to master.

Testing Expectations#

Before opening a PR:

  • Run formatters/linters relevant to changed files. For DUMB backend changes, run make format before make verify; make verify is check-only and does not rewrite Black formatting drift.
  • Run focused tests for your change.
  • Include manual verification notes when automated tests are not available.

Service update regression suite#

DUMB retains a disposable previous-version-to-latest regression matrix for service installers and updaters. It runs selected services in isolated Docker containers with fresh configuration, data, logs, mounts, and databases; it never downgrades the operator's live service data.

From the DUMB repository on a Docker-capable host:

make regression-image
make regression-service-updates REGRESSION_JOBS=2

Two workers are the conservative default because source builds can be CPU-, memory-, and disk-intensive. Cases run concurrently when they do not share a worker, while each service's install, update, and health checks remain ordered within that container. Use REGRESSION_ARGS="--case <case-id>" for focused coverage or REGRESSION_ARGS="--include-pending --case <case-id>" while qualifying a new matrix entry. REGRESSION_ARGS="--use-gh-auth-token" uses the current GitHub CLI token without printing it and avoids anonymous API limits.

The matrix lives in scripts/service_update_regression_matrix.json. Qualified cases run by default; pending entries document services that still need deterministic prior versions, dependency fixtures, or disposable credentials. Per-case logs and an aggregate JSON report are written under the ignored .regression-reports/ directory. The verified install cache is retained separately so repeat runs also exercise safe cache reuse.

Docs Contributions#

Docs live in DUMB_docs.

  • Update docs when API behavior, onboarding flow, or service settings change.
  • Keep examples and screenshots current.
  • Prefer small, reviewable docs PRs.

Non-Code Contributions#

Even if you don't write code, there are plenty of ways to help!

Star the Repo#

A quick way to show support is by starring the project on GitHub: github.com/I-am-PUID-0/DUMB

Join the Community#

  • Ask questions or help others in the DUMB Discord server
  • Boost the server if you find it helpful
  • Share your setups and improvements

Help with Documentation#

  • Found a typo or confusing section in the docs? Open an issue or pull request.
  • Suggestions and feedback on existing documentation are always welcome.

Monetary Contributions?#

Roadmap & Feedback#

If you have ideas, feature requests, or feedback:

  • Create a GitHub Issue

  • Discuss in the community before starting larger PRs