Release 0.5

[bernhardmgruber]

This issue tracks all missing steps for the LLAMA 0.5 release

• All examples must build.
• All unit tests must build and pass.
• The documentation must build (check readthedocs).
• Verify that the documentation is up-to-date.
• Verify that the README.md is up-to-date.
• Verify that the CHANGELOG.md is up-to-date.
• Replace .zenodo.json by CITATION.cff: Replace zenodo.json by CITATION.cff #696
• Verify that .zenodo.json is up-to-date, especially regarding contributors.
• Bump the version number in llama.hpp and docs/conf.py: Bump LLAMA version to 0.5 #695
• Test in downstream projects (AdePT).
• Test in downstream projects (picongpu). Works with nvcc 11.8, but not with 12.0. It is not clear whether the error for the latter stems from LLAMA or the glue code for the integration :/
• Merge outstanding PR: Improve and rename recordCoordTags into prettyRecordCoord #693
• Merge outstanding PR: Fix some ODR violations #698
• Set release date in CHANGELOG.md and commit
• Draft a new release on GitHub: https://github.com/alpaka-group/llama/releases.
• Release

[bernhardmgruber]

Released: https://github.com/alpaka-group/llama/releases/tag/v0.5