I'm also eager to start using new feature 😁

I would maybe keep it at least until its EOL (https://endoflife.date/python, not sure if it's the most accurate prediction but https://www.python.org/dev/peps/pep-0494/#lifespan suggests something like that). Do you mean deprecate and just emit a warning or drop completely? As a library I'd keep it a little conservative, I know Ansible can be like that and the community.general collection depends on this for the gitlab modules.

But maybe that shouldn't stop us from adding from __future__ import annotations. We could add https://pypi.org/project/future-annotations/ to requires for python<3.7, it's only a small penalty for 3.6 users at install time.

I guess by "deprecate" I mean stop supporting. As in drop-completely. Would make sense to me for the 3.0 release. The 2.X release would still have support for Python 3.6.

Looks like Python 3.6 is End-of-Life in less than two months... To me that is pretty much EOL to me.

Good point. We could maybe do it in December with 3.0.0.

Yeah that makes sense. Python 3.6 is officially supported until the 23th of December, according to https://endoflife.date/python

Cool. So if we have made a decision to drop Python 3.6 support for the 3.0.0 release, we probably could start the work now in the main branch.

What do you think?