Project Management¶
The people working on PUDL are distributed all over North America. Collaboration takes place online. We track our work in a public GitHub Project so you can follow our progress.
Issues and Project Tracking¶
We use Github issues to track bugs, enhancements, support requests, and just about any other work that goes into the project. Try to make sure that issues have informative tags so we can find them easily.
Our GitHub Workflow¶
We have 3 persistent branches:
main
(the default branch),nightly
, andstable
.We create temporary feature branches off of
main
and make pull requests intomain
throughout our 2 week long sprints. All code that’s merged intomain
should have passed our CI tests and been reviewed by at least one other person.Every night the
main
branch is used to run the Nightly Data Builds. If the builds are successful, then thenightly
branch is automatically updated to point to the latest commit onmain
. If the builds fail, then thenightly
branch is left unchanged.Every time we do a versioned data release, the
stable
branch is updated to point to the commit associated with the most recent release.
Pull Requests¶
Before making a PR, make sure the tests run and pass locally, including the code linters and pre-commit hooks. See Set Up Code Linting for details.
Don’t forget to merge any new commits to the
main
branch into your feature branch before making a PR.If for some reason the continuous integration tests fail for your PR, try and figure out why and fix it, or ask for help. If the tests fail, we don’t want to merge it into
main
. You can see the status of the CI builds in the GitHub Actions for the PUDL repo.Please don’t decrease the overall test coverage – if you introduce new code, it also needs to be exercised by the tests. See Testing PUDL for details.
Write good docstrings using the Google format
Pull Requests should update the documentation to reflect changes to the code, especially if they change something user-facing, like how one of the command line scripts works.
Releases¶
The PUDL data processing pipeline isn’t intended to be used as a library that other Python packages depend on. Rather, it’s an end-use application that produces data which other applications and analyses can consume. Because of this, we no longer release installable packages on PyPI or
conda-forge
.Periodically, we tag a versioned release on
main
using a calendar based version, likev2023.07.15
. This triggers a snapshot of the repository being archived on Zenodo.The nightly build outputs associated with any tagged release will also get archived on Zenodo here and be made available longer term in the AWS Open Data Registry.
User Support¶
We don’t (yet) have funding to do user support, so it’s currently all community and volunteer based. In order to ensure that others can find the answers to questions that have already been asked, we try to do all support in public using Github Discussions.