How datahub-project/rfcs works
Main Components
• RFC Process Repository – The overall container for all RFC-related documents.
• Active RFCs – This is the “active” directory, which holds subfolders for various RFC proposals (e.g., “1778-dashboards”, “1812-ml_models”, etc.).
• Finished RFCs – The “finished” directory, which appears to manage RFCs that have been concluded (e.g., the presence of a .lock file may indicate state/version control).
• Templates – The “templates” directory containing template files used to start RFCs (e.g., “000-template.md”).
• Supplemental Components – Other directories like “access-control” and “tags” should also be recognized as specialized RFC topics.
Understand the Repository Purpose and Flow
• The README points to DataHub’s online RFC process. This gives context that the repository is not for running production code but for managing RFC proposals.
• Think of the repository as a workflow system: starting with a template, moving into an “active” phase for discussion and development, and finally moving to “finished” status after review and approval.
Data Flow and Lifecycle
• Start at Templates: Proposals are created from a template.
• Move to Active RFCs: The RFC enters a discussion/development phase.
• End at Finished RFCs: Once approved, the RFC (or the decision output) is marked as finished.
Architectural and Design Patterns
• Modular Organization: Each RFC proposal is kept independent as a folder under “active” or “finished” which promotes separation of concerns.
• Lifecycle Management: The clear transition from template → active → finished is an architectural pattern for managing state.
• Extensibility: The repository structure can be extended with new RFCs by simply adding new directories under “active” (or moving them to “finished” post-approval).