You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
RUN micromamba install -c conda-forge -y python=3.8 \
as the images are using micromamba (great!) can these be updated to use Python 3.10? It would be great to have a more modern CPython runtime that is going to be more robust to things like NEP 29. While the experiment ecosystems lag considerably behind even Python's EOL schedule because of lock in to CentOS 7 support
NEP 29 is taken seriously in the Scientific Python community (c.f. SPEC 0) and given ongoing efforts to have better integration, support, and interaction between Scientific Python and Scikit-HEP it should be assumed that Awkward and boost-histogram (at least) will probably be following SPEC 0.
Unless there is a known bug that keeps software from working on Python 3.10 I would highly recommend having it be the default (here I'm currently not advocating for bumping to Python 3.11 (for the time being) as TensorFlow v2.11.0 still doesn't support it).
The text was updated successfully, but these errors were encountered:
At the moment the Python runtime version used for the
cento7
"base" images is Python 3.8ML-Containers/centos7/base.Dockerfile
Line 33 in b4ca7cc
as the images are using
micromamba
(great!) can these be updated to use Python 3.10? It would be great to have a more modern CPython runtime that is going to be more robust to things like NEP 29. While the experiment ecosystems lag considerably behind even Python's EOL schedule because of lock in to CentOS 7 supportNEP 29 is taken seriously in the Scientific Python community (c.f. SPEC 0) and given ongoing efforts to have better integration, support, and interaction between Scientific Python and Scikit-HEP it should be assumed that
Awkward
andboost-histogram
(at least) will probably be following SPEC 0.Unless there is a known bug that keeps software from working on Python 3.10 I would highly recommend having it be the default (here I'm currently not advocating for bumping to Python 3.11 (for the time being) as TensorFlow
v2.11.0
still doesn't support it).The text was updated successfully, but these errors were encountered: