How do people organize open source collaboration at scale?

This FOSS Foundations directory is collection of datasets about the many excellent non-profit Foundations that directly host or otherwise help open source projects succeed.

We welcome academic researchers studying the organizations behind open source, and serve as a showcase for the organizations and communities behind much of the software ecosystem today.

Goals

  • Create a rich metadata directory about nonprofit open source supporting organizations, including financial, sponsorship, governance, leadership, and other metadata.
  • Store data in easy-to-update formats, and provide reliable, faceted, and discoverable access.
  • Curate accurate and historical data with discoverable attestations.
  • Keep the site and tooling as zero/low maintenance as practical.
  • Have a simple PR-driven data submission process that prevents errors.
  • Attract multiple contributors to create community governance.
  • Build a resource for the FOSS ecosystem that will last for 50+ years.

Non-Goals

Volunteers are welcomed to help turn non-goals into goals, if there is interest and energy to build them!

  • Tracking commercial entities.
  • Tracking non-open source or free software related entities.
  • Tracking non-entity projects; i.e. open source projects without legal entities.
  • Building datasets easily accessible elsewhere, like Ecosyste.ms.
  • Fancy features, UI, or hard to maintain visualizations. We provide a data store that others can repurpose.

Data

Read more about our data philosophy.

  • Store structured metadata about Foundations in _foundations/asf.md frontmatter, parseable as YAML/JSON.
    • One file per foundation makes PRs simple to manage and associate with a responsible contributor.
    • Frontmatter metadata can be exposed by OpenAPI for programmatic access.
    • Jekyll turns foundation data and Markdown bodies to human-readable websites automatically.
    • OpenAPI and GitHub Actions tooling provides linting to keep data formats clean.
  • Using flat files and GitHub Pages means zero maintenance and detailed change history.
  • Markdown body can be used for additional, unstructured data or descriptions.
  • Use common human-findable identifier strings to tie all models together.
  • Create good-enough data structures for practitioners and academics alike.

Research

We are looking for research collaborations, both to validate that the metadata we’re tracking is valuable to researchers, and to seek help in expanding our datasets. Many of the research topics related to this site will result in new paper topics, as well as impactful results for the communities you’re studying.