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
I think we should drop the markers for 3.5 and probably 3.6, and just raise python_requires appropriately. Those on old python versions will still get functioning oldest-supported-numpy packages, and almost by definition, those old language versions will not receive any new architecture support. However, they still have a maintenance cost, as seen in cases like #71, so I think we should drop them.
Python 3.7 is also EOL this month, but it's a borderline case.
The text was updated successfully, but these errors were encountered:
I think we should drop the markers for 3.5 and probably 3.6, and just raise
python_requires
appropriately. Those on old python versions will still get functioningoldest-supported-numpy
packages, and almost by definition, those old language versions will not receive any new architecture support. However, they still have a maintenance cost, as seen in cases like #71, so I think we should drop them.Python 3.7 is also EOL this month, but it's a borderline case.
The text was updated successfully, but these errors were encountered: