For Python 3.12, the numpy>=1.26.0b1 means that using pip's default upgrade strategy, this will also pick up release candidates, which might not be desirable (c.f. https://github.com/scikit-hep/cabinetry/pull/477)
but there hasn't been a release since then. A new release would resolve this issue, without users needing to explicitly guard against release candidates.
Describe the bug
This might not really be a "bug", but it is different from the expected user experience.
hist
v2.7.2
has different NumPy requirements between Python 3.12 and older Pythonshttps://github.com/scikit-hep/hist/blob/8ab95946809e8195fcbecc7cc398738af1d2267d/pyproject.toml#L41-L42
For Python 3.12, the
numpy>=1.26.0b1
means that usingpip
's default upgrade strategy, this will also pick up release candidates, which might not be desirable (c.f. https://github.com/scikit-hep/cabinetry/pull/477)This was fixed in https://github.com/scikit-hep/hist/pull/543 with the move to
https://github.com/scikit-hep/hist/blob/9c3a275ff3f614d5ea201149b7851905a4c2e8b4/pyproject.toml#L41
but there hasn't been a release since then. A new release would resolve this issue, without users needing to explicitly guard against release candidates.
Steps to reproduce