Skip to content

Latest commit

 

History

History
33 lines (25 loc) · 4.24 KB

CONTRIBUTING.md

File metadata and controls

33 lines (25 loc) · 4.24 KB

Contributing to Science on schema.org

The goal of science-on-schema.org is to foster community agreement on guidelines for using schema.org for science-related resource descriptions. Thus, we welcome and encourage community contributions to the work, including proposals for new guidelines, clarifications on the existing guidelines, and proposals for extensions to the schema.org model to accomodate new fields. Thus, community members are invited to contribute through:

  • Creating and discussing proposals through GitHub issues
  • Implementing proposed features by editing the guidelines and extension documentation, and editing the Architectural Decision Record for the proposal
  • Issuing a Pull Request to request review of the proposed changes, and to discuss the feature and iterate on development

Development and Release process

Because people will review our guidance documents online, it will be helpful to be explicit about versioning and releases of the guidance and other documents. In particular, in GitHub, users will normally see the main branch, which therefore should reflect the current stable release of the documentation (rather than confusing people with in-progress proposed changes that are not yet released).

Consequently, we will use a GitFlow-inspired release model in which the main branch always reflects the current stable release, a develop branch is used for merging finished proposals being prepared for release, and feature branches are used for creating changes to implement specific proposals that are reflected in an Architectural Decision Record (ADR). For changes that do not require a formal decision via an ADR, such as spelling corrections, grammatical rewording, etc., maintainers can commit changes directly to the develop branch, and other contributors can do a pull request directly against the develop branch. The use of feature branches is really focused on managing proposals that need discussion, review, and a decision through an ADR. Maintainers will make judgement calls on whether an ADR is needed, and might convert contributed pull requests to a feature branch if they determine that an ADR is needed.

In particular:

  • The main branch of the GitHub repository always reflects the most current release
  • A develop branch is used for development work to extend the guidelines for each of the ADRs and for other more minor changes
  • Each ADR is associated with a pull request against the develop branch, and the PR should reference the issue number for the ADR in the template
    • Any PRs should be merged into the develop branch when they have been approved as a good implementation of the ADR
  • The develop branch is merged into main and tagged to create a new release when it has been approved for release
  • Pull requests should be based on a separate feature branch that is appropriately named for the feature and include the issue # that is being addressed for the feature (e.g., feature_30_release_workflow)
  • Discussion of the proposed ADR should occur in the associated GitHub issue
  • When agreement has been reached, the proposed changes and the ADR should be updated in the feature branch, and then merged into the develop branch.
  • When the set of features targeted for a release are complete and review has been finished, the develop branch will be merged into the main branch and tagged as a release.
  • Release tags will follow semantic versioning, and should be used to create an associated release on GitHub that allows files to be downloaded.

Consequences

  • People can easily follow the current stable release guidelines on the GitHub page
  • The community can easily propose changes through pull requests against the development branch
  • There are clear linkages between github issues, the ADR decision process, and the release process
  • Development for core committers is somewhat more complicated as people need to understand and use the branching model, but this doesn't affect third-party contributors
  • Some contributors might accidentally submit pull requests against the main branch, which will need to be retargeted by a maintainer against the develop branch before it can be merged