Christian, what is the status of PEP 543? Is someone actively working on it?  Should we hope for a resolution (i.e. an implementation accepted in the stdlib) in the near future?

I'm worried about such purity arguments.  It reminds me of when we decided to freeze distutils because distutils2 was to become the future (and then distlib took that place, etc.).  In the end "the future" hasn't really happened and distutils has rotten without a clear path forward for users (except the venerable setuptools, which has its own issues).