You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently, the v3-dev branch is failing the check-release workflow. The branch deviates from main at v2.26.0, and now that v2.28.0 exists, the workflow fails with the message:
ValueError: Tag v2.28.0 already exists
...
subprocess.CalledProcessError: Command 'jupyter-releaser draft-changelog' returned non-zero exit status 1.
Proposed Solution
This workflow can be fixed by updating the version of the v3-dev branch manually to some v3 version. We should decide how the first release of v3-dev will be versioned, since we will have to do this later anyways.
Using a pre-release will allow us to make breaking changes as needed when uncovering issues during usability testing. Therefore, it seems preferable to use a pre-release version for the first v3-dev release.
I recommend using the version 3.0.0-alpha.0 as the version of the first v3 release performed by Jupyter Releaser. This should be distributed as 3.0.0a0 on PyPI. This also seems aligned with the convention used by JupyterLab.
Contributors, please leave feedback in this issue.
I recommend using the version 3.0.0-alpha.0 as the version of the first v3 release performed by Jupyter Releaser. This should be distributed as 3.0.0a0 on PyPI. This also seems aligned with the convention used by JupyterLab.
Problem
Currently, the
v3-dev
branch is failing thecheck-release
workflow. The branch deviates frommain
at v2.26.0, and now that v2.28.0 exists, the workflow fails with the message:Proposed Solution
This workflow can be fixed by updating the version of the
v3-dev
branch manually to some v3 version. We should decide how the first release ofv3-dev
will be versioned, since we will have to do this later anyways.Using a pre-release will allow us to make breaking changes as needed when uncovering issues during usability testing. Therefore, it seems preferable to use a pre-release version for the first
v3-dev
release.I recommend using the version
3.0.0-alpha.0
as the version of the first v3 release performed by Jupyter Releaser. This should be distributed as3.0.0a0
on PyPI. This also seems aligned with the convention used by JupyterLab.Contributors, please leave feedback in this issue.
Additional context
We have cut pre-releases of Jupyter AI in the past: https://github.com/jupyterlab/jupyter-ai/releases/tag/v2.10.0beta0
The text was updated successfully, but these errors were encountered: