We saw it at #194 and other PRs. Sometimes the tests fail on one machine and pass on others, with the different results on the second or third digit for some values. This is probably due to different versions of AVL, Python, or even dependencies.
This should be solved so we have consistency across machines and Travis.
We saw it at #194 and other PRs. Sometimes the tests fail on one machine and pass on others, with the different results on the second or third digit for some values. This is probably due to different versions of AVL, Python, or even dependencies.
This should be solved so we have consistency across machines and Travis.