Skip to content

Migration FAQ

Jump to bottom
Adrian Edwards edited this page Apr 30, 2026 · 2 revisions

Why is there a new project that I need to migrate to?

See this announcement from the CHAOSS board.

Why was the new name chosen?

The name CollectOSS (pronounced: collect us) was picked as a nod to the purpose of the project (collecting data on open source repositories) while being easy-to-remember. The OSS suffix is also intended to evoke a shared purpose with the CHAOSS organization, whose name also ends with OSS.

Who are the maintainers?

To provide the project with a strong community foundation, this project is stewarded by a transition team of maintainers. Those maintainers are:

  • Adrian Edwards
  • Josh Berkus
  • Cali Dolfi
  • Shlok Gilda

More information can be found in MAINTAINERS.md

How will the project be governed?

The transitional maintainer team has introduced a governance process that is based on a template from the CNCF. This document is available at GOVERNANCE.md in the repository and outlines the policies and procedures for maintainer addition, maintainer removal, voting, governance changes, and more. Feedback is always welcome and changes to this document are possible with a ⅔ vote of the maintainers.

Why did the repository get rebased?

The disruption of our original repository offered us a rare opportunity to review and clean up up some deeply-nested bloat within the repository history that would otherwise have been too disruptive to clear out.

Throughout the nearly 10 years of commit history in the project, a lot of extraneous files were added to version control and later deleted. These files included:

  • build output generated from the source such as javascript builds of the frontend and docs, as well as source maps and a node_modules folder
  • *.whl and *.egg and other python packaging files that can be regenerated
  • *.ipynb - since-deleted jupyter notebooks which often contain lots of generated output. These were likely migrated to https://github.com/chaoss/wg-data-science/ a long time ago.
  • Other large since-deleted binary files (such as fonts)
  • older editor-specific configurations and files that have since been added to .gitignore
  • large data files that appear to be for sample data
  • *.model - pre-trained ML models that have since been deleted. These took up a lot of space and were removed in 2022 "until they are completely tested".

To make these changes, a rebase of the entire repository history was required.

This change has resulted in the cloned size of the repository dropping by nearly 50% (~118 MiB vs 220 MiB). Removing this old baggage means a more efficient repository that prioritizes the forward movement of the project and friendliness to new contributors over the preservation of every historical artifact.

This rebase has not modified any authorship attribution, timestamps, or licenses that were part of the code. All code retains its original authorship attribution under the MIT license.

If you have questions or would like help understanding how this change affects you, please reach out on slack and the maintainers are happy to assist you!

What about the issues?

Because GitHub doesn’t offer a suitable issue-copying feature, we have copied over select issues from the original repository in order to preserve issues that:

  • Were currently being worked on
  • Are still valid issues with the codebase
  • Are still useful to refer to for historical reference

This was done using a custom-built github bot user that uses the Github API to copy over both issues and their comments. This bot also includes a small section to give credit to the original issue and author since the Github API does not allow bots to post on behalf of other users or set timestamps to a date in the past.

If you notice any problems with the issues that were copied over, please reach out on slack and the maintainers are happy to assist you!

What else changed?

As part of this renaming process, we noticed many places in the project files and documentation that refer to older, unsupported methods of installing the project. Many of these references were removed to help avoid confusion and make the transition process easy for both new and existing contributors.

Because this is a new project, the version numbering scheme for the project has also been reset, with this initial release being tagged as v1.0.0. This version numbering logically follows from the v0.X.Y numbering scheme that contributors are likely used to, while signalling that this is the start of something that is both new and stable.

Clone this wiki locally