Upgrading Python

We officially support the 4 oldest, still maintained Python versions. This means that the newest Python version(s) may not be tested, but the expectation is that those should also work after enough time has passed for the ecosystem to adapt.

TODOs when upgrading Python

When a Python version is end of life (EOL):

• remove said version from tests.yml, add the next Python version not yet being tested (e.g. remove 3.11, add 3.15)
• further workflows that might need updates: build_{pr_}documentation.yml, integration_tests.yml, triage_prs.yml, etc. (check all workflows that use Python)
• setup.py: bump min Python version
• install.md
• pyproject.toml (may require updating ruff rules afterwards)
• if a useful new feature is enabled by dropping the old Python version, consider using it
• post in Slack channel, as dependencies may need upgrading too