v0.14.1 point release

[sk1p]

Issues/PRs that should be included:

• Fix compatibility with numpy>=2.0.0 #1656
• Docs update for point release #1660

@uellue @matbryan52 would any of you like to take charge of this release?

Release checklist

Issues and pull requests to be considered for this release:

• Fix compatibility with numpy>=2.0.0 #1656

Step 0: before cutting the release candidate

• Document point release with version downgrade, while we are at it.
• Review open issues and pull requests
• Confirm that pull requests and issues are handled as intended, i.e. milestoned and merged
  in appropriate branch.
• [N/A] License review: no import of GPL code from MIT code
  pydeps --only "libertem" --show-deps --noshow src\libertem | python scripts\licensecheck.py
• Run full CI pipeline, including slow tests, on Azure DevOps and run the Thorough workflow on GitHub Actions
• [N/A] Handle deprecation, search the code base for DeprecationWarning
  that are supposed to be removed in that release.
• [N/A] GUI dependency update with npm install
• [N/A] Review https://github.com/LiberTEM/LiberTEM/security/dependabot and update dependencies
• [N/A] Full documentation review and update, including link check using
  sphinx-build -b linkcheck "docs/source" "docs/build/html"
• Update the expected version in notes on changes, i.e. from 0.3.0.dev0
  to 0.3.0 when releasing version 0.3.0.
• Update and review change log in docs/source/changelog.rst, merging
  snippets in docs/source/changelog/*/ as appropriate.
• [N/A] Edit pytest.ini to exclude flaky tests temporarily from release
  builds, if there are currently any flaky tests
• [N/A] Update the JSON files in the packaging/ folder with author and project information

Step 1

• Create a release candidate using scripts/release. See
  scripts/release --help for details. Example command:
  ./scripts/release bump v0.3.0rc1 --tag, then push to GitHub
• Confirm that wheel, tar.gz, and AppImage are built for the release candidate on
  GitHub
• Confirm that a new version with the most recent release candidate is created on
  Zenodo.org that is ready for submission.

Step 2: using the release candidate package

• [fix any fallout after point release] Run tests for related packages w/ new LiberTEM version
  • LiberTEM-live
  • LiberTEM-holo
  • LiberTEM-blobfinder
  • ptychography40
  • LiberTEM-iCoM
• Run complete test suite, including slow tests that are deactivated by default
  and tests that require sample files or CUDA support.
• Install release candidate packages in a clean environment
  (for example:
  python -m pip install 'libertem==0.2.0rc11')
• Test the release candidate docker image
  • Confirm rc images and tags on https://ghcr.io/libertem/libertem
• Quick GUI QA: open in an incognito window to start from a clean slate
  • Correct version info displayed in info dialogue?
  • Run a couple of analyses to verify it is not entirely broken

Step 3: bump version and let release pipeline run

• Final version bump: ./scripts/release bump v0.3.0 --tag, push to github
• After pipeline finishes, write minimal release notes for the release and publish the GitHub release

Step 4: after releasing on GitHub

• Confirm that all release packages are built and release notes are up-to-date
• Install release package
• Confirm correct version info
• Confirm package upload to PyPI
• Confirm images and tags on https://ghcr.io/libertem/libertem
• Publish new version on zenodo.org
• Update documentation with new links, if necessary
  • Add zenodo badge for the new release to Changelog page
• Conda packaging: review PRs on https://github.com/conda-forge/libertem-feedstock/pulls
• Send announcement message on mailing list
• Edit pytest.ini to include flaky tests again
• Bump version in master branch to next .dev0 (./scripts/release bump v0.X.0.dev0 --commit)
• Add to institutional publication databases
• [skip] Add the current LiberTEM version to CVL - add both the singularity and the .desktop file!

[uellue]

@sk1p sorry to bother, but what were the steps for a point release again? Sth about setting the version by hand?

We didn't merge anything of importance since the last release, so no need for a stable branch.

[sk1p]

@sk1p sorry to bother, but what were the steps for a point release again? Sth about setting the version by hand?

Yup, either setting the version to 0.14.1.dev0 by hand, or using the release script to "bump backwards" with --force (IIRC)

We didn't merge anything of importance since the last release, so no need for a stable branch.

👍

[matbryan52]

@uellue @matbryan52 would any of you like to take charge of this release?

I am happy to test and help but I haven't ever been close to the active part of release cutting before, would be coming from a ways behind. If @uellue you want to look at it together I am there.

[uellue]

Yes, --force made it downgrade. Thx! Now waiting for the thorough tests.

[uellue]

Cluster analysis doesn't like NumPy 2 on 3.9: https://github.com/LiberTEM/LiberTEM/actions/runs/9567900368/job/26376754929

--> #1658

[matbryan52]

Cluster analysis doesn't like NumPy 2 on 3.9: https://github.com/LiberTEM/LiberTEM/actions/runs/9567900368/job/26376754929

--> #1658

Looks like the scikit-image release fixed the cluster analysis problem on py39 : https://github.com/LiberTEM/LiberTEM/actions/runs/9567900368/job/26404990589

[uellue]

I also confirmed that LiberTEM still works with NumPy < 2, since we recommend it. In a call until 11:00, then I will do short docs update.

[matbryan52]

Just a warning for now, I imagine this predates numpy2.

src\libertem\io\dataset\npy.py:11
  ~\Workspace\libertem_fork\src\libertem\io\dataset\npy.py:11: DeprecationWarning: `np.compat`, which was used during the Python 2 to 3 transition, is deprecated since 1.26.0, and will be removed
    from numpy.compat import long, asstr

[uellue]

Just a warning for now, I imagine this predates numpy2.

src\libertem\io\dataset\npy.py:11
  ~\Workspace\libertem_fork\src\libertem\io\dataset\npy.py:11: DeprecationWarning: `np.compat`, which was used during the Python 2 to 3 transition, is deprecated since 1.26.0, and will be removed
    from numpy.compat import long, asstr

I'm preparing a PR to integrate the upstream changes which fix this. :-)