New production version, likely 0.12.0

[pc494]

With #469 coming in and fixing an important bug with significant downstream (--> diffsims --> pyxem) implication it seems likely we will need a new release of orix. Given what's currently in the changelog this is likely to be a minor release. I'll use this issue to collate what we may (or may not) need to get in before making the release.

Tagged PRs in unclear states

#442
#451
#457
#460

Milestone issues

#439

Outline of plan (in planned order of merges)

• CI: Add pre-commit CI #472
• Ensure atom positions are expressed in the new basis when initializing Phase with structure #469 review, adjust and merge
• copyright Documentation: copyright from 2023-->2024 #473
• Add Sampling #460 which helps with diffsims a fair bit
• deps

I think the others can be left for now.

[pc494]

For what it's worth, I would be happy to merge #469, deal with the copyright notice and send it to production. The bug being solved by #469 is blocking diffsims and pyxem at the moment.

[CSSFrancis]

@pc494 #460 Just needs a review. I think that one should be good to go.

We should do #472 Just to help us get some of the other things in. I can help review but can't directly clean up others work to help get things merged which is likely the path we would have to take to get this released quickly.

More smaller releases is good.

[hakonanes]

I agree that a swift minor release is good. I'll have a look at the docs.

[pc494]

Is it worth putting out a release candidate while we wait for some progress on #475?

[CSSFrancis]

@pc494 or @hakonanes Do you want to make a release? I can do it as well if you would like.

[hakonanes]

If it's OK with you I'd like to do it this evening after looking over the latest changes.

[pc494]

If it's OK with you I'd like to do it this evening after looking over the latest changes.

This works for me :)

[hakonanes]

I'll have to look over the docs before release, which I'll do tomorrow evening. Hence, I can make the release then at the earliest.

If you would like to give feedback on the changes, we should wait until Thursday.

[pc494]

#478 looks good so far. I could look over changes Thursday morning UK time, would it be okay @hakonanes for you to set stuff up this evening and then (assuming all goes well) make a release on Thursday evening?

[hakonanes]

Assuming all checks pass, #478 is done. It just needs a review. I'm OK with releasing after that.

What do you mean by the TODO called "deps" in the top comment, @pc494? Is it a check for minimal supported versions?

[hakonanes]

Regarding minimal supported versions (Python 3.7), see #482.

[CSSFrancis]

#483 Should be included as well to support numpy 2.0.0

[hakonanes]

I've opened #487 to bring all changes since the last release onto the main branch for tagging and releasing.

[hakonanes]

It's on PyPI: https://pypi.org/project/orix/.

Waiting for conda-forge. This evening or tomorrow morning.

[hakonanes]

Other updates:

• Zenodo: https://zenodo.org/records/10966629
• New stable docs: https://orix.readthedocs.io/en/stable/index.html?

[hakonanes]

For some unexplicable reason, our build action

orix/.github/workflows/python-publish.yml

Lines 39 to 41 in 8620292

	- name: Build package
	run: |
	python -m build

created a distribution file with name orix-0.12.tar.gz, without the patch number. It has always included that before. The version is listed and interpreted correctly all other places I've looked.

Our conda-forge feedstock build therefore failes, because it looks for orix-0.12.0.tar.gz and can't find it. I had to add a temporary_version parameter to the meta.yaml script to temporarily point conda-forge to the orix-0.12.tar.gz file. This must be removed when releasing a new version!

[hakonanes]

It's on conda-forge: https://anaconda.org/conda-forge/orix.