Thanks for your interest in applying for Google Summer of Code with conda-forge. We welcome applications from individuals from all backgrounds, identities and abilities and encourage applications individuals from under-represented groups in tech.
conda-forge is a community effort and a GitHub organization which contains repositories of conda recipes and thus provides conda packages for a wide range of software. The built distributions are uploaded to anaconda.org/conda-forge and can be installed with conda.
2023 marks the 8th anniversary of the conda-forge organization. Over these 8 years it has served more than 1.2M package artifacts, which account for 10B downloads via anaconda.org. None of this would be possible without the work of more than 4.6K volunteers!
conda-forge is built for and around the conda packaging ecosystem. A conda recipe contains the dependency metadata and instructions to build and package a particular project, usually from source.
New recipes are first submitted to the conda-forge/staged-recipes
repository via a pull request (PR). Once this is reviewed and approved, the recipe is merged and granted its own repository, called feedstock. A feedstock is a standard GitHub repository within the conda-forge organization, containing the user-provided recipe/
and the supporting configuration and tooling required for the builds.
For each PR merged in a feedstock, a series of artifacts is built for the package (changes such as releasing a new version or adding a new dependency require rebuilding the package for ecosystem-wide compatibility). All contributions to a feedstock happen through PRs.
The conda-forge organization is led by the core team. The core team also receives support from many volunteers, like staged-recipes reviewers or the domain-specific help-*
teams. Each feedstock is mainly maintained by its team of collaborators named after it. Feedstock teams are fluid by design: anyone can step down at any moment and join and take over as long as the feedstock team accepts the PR-driven application.
conda-forge is an open-source project; it is built in the open for traceability and transparency. Continuous integration (CI) logs are publicly available, and everyone is welcome to contribute. We follow a strict fork-based Github-flow workflow: the feedstock branches will build and upload the artifacts to the validation server. For this reason, pull requests must always be opened from personal forks. This workflow is also followed in conda-forge repositories that are not responsible for producing packages.
conda-forge.org was created almost 8 years ago (when the whole conda-forge project was started). After 8 years, the site has become outdated. For example, it has not been optimized for mobile use. Moreover, it is not WCAG (Web Content Accessibility Guidelines) conformant, which makes it inaccessible to many users, especially those with disabilities or those that rely on assistive technology to interact with the web.
This project aims to deliver a new website for conda-forge.org that meets WCAG accessibility standards and improves performance while providing an excellent contributor experience.
The current Sphinx-based conda-forge.org website is mainly built from the conda-forge.github.io
repository and extended by status
- which generates a basic conda-forge
status dashboard - and a conda-forge blog
.
A Docusaurus-based prototype website is available at cf-infra-docs.netlify.app, where new documentation is being written for the conda-forge infrastructure. We will use this site as the playground for the website modernization.