It looks like the failed runs are due to int.as_integer_ratio not being available in CPython 3.6. Should I just delete the tests?

Thanks for your contributing.
Upgrading test to use 3.8 will be the best. If it is hard, add an if-statement for python version like:

import sys
if sys.version_info >= (3, 8):
    # run 3.8 test here

I'll try upgrading the Python version in GitHub workflow first.

Upgrading test to use 3.8 will be the best. If it is hard, add an if-statement for python version like:

@youknowone have already tried upgrading to 3.8 in #1705. I think that commenting out the tests and adding a TODO will be good enough.

@palaviv, sure, I'll modify this PR.

Thank you for contributing!