We are already running python3 in production and python2 support is ending. We can drop the python2 compatibility layer (from isort for example), and also don't run python2 on Travis, instead start testing with python 3.7
I'm +1 on this. The corporate side is still Python 2, so that is the primary blocker currently, but I'm +1 on it after that.
Just linking to some tools that can help with the cleanup
I'm a definite +1 on dropping support for Python2. Supporting both comes at a cost. I have a few notes:
v3.0.0).six and future (we use both) and we'll slowly clean up code that uses them.Agreed on lumping this into v3.0, moving discussion here to our on-going discussion on deprecation #4594
Nevermind, breaking out a discrete ticket! We'll address this at v3 at the end of the year.
I opened https://github.com/rtfd/common/pull/19 and https://github.com/rtfd/common/pull/20 to help with the removal of the python2 layer
Most helpful comment
I'm a definite +1 on dropping support for Python2. Supporting both comes at a cost. I have a few notes:
v3.0.0).sixandfuture(we use both) and we'll slowly clean up code that uses them.