On Thu, Apr 15, 2021 at 12:01 PM Samuel Colvin <samcolvin@gmail.com> wrote:
I've read the recent discussions <https://mail.python.org/archives/list/python-dev@python.org/thread/QSASX6PZ3LIIFIANHQQFS752BJYFUFPY/#UITB2A657TAINAGWGRD6GCKWFC5PEBIZ> regarding PEP 649 and PEP 563 with interest, Larry Hastings recently contacted me when prompted <https://mail.python.org/archives/list/python-dev@python.org/message/YKVYJMLUWUVT4KMLUNEQYVBZWNAPR4GV/> to do so in a related discussion.
I maintain pydantic <https://pydantic-docs.helpmanual.io/> which uses type annotations to provide runtime data validation.
I'm extremely worried that if PEP 649 gets rejected, pydantic will not be able to fully support python 3.10 and may even end up getting abandoned, at best it will be much slower and more brittle.
Please, please put pragmatism over principle and accept PEP 649.
Please don't phrase the decision in these terms. While for Pydantic PEP 649 is more pragmatic, that does not mean PEP 563 isn't pragmatic for other people for other reasons. Making this an "us versus them" discussion just makes the whole situation feel confrontational when instead everyone is trying to figure out the best thing for everybody when there's no perfect answer.