Yes -- I believe we agreed on the "best effort" approach, but I never got around to updating the actual PEP.

I made a PR just now though: https://github.com/python/peps/pull/995. @Mark do you mind taking a look when you have time?

Thanks! (And sorry for the delay)
-- Michael

On Mon, Apr 15, 2019 at 11:39 AM Guido van Rossum <guido@python.org> wrote:
Did this discussion get resolved? The conclusion seems to be that the PEP should specify best-effort for this case? Can one of you submit a PR to update the PEP (and then the other would review it before I merge it).

On Mon, Mar 18, 2019 at 10:26 AM Mark Mendoza <mendoza.mark.a@gmail.com> wrote:
Totally agree re: allowing Literals to interact normally with the type system as much as possible.  I think option 1 (not allowing Literals in "implicit" type variables) is unfortunately a violation of that principle and could get confusing.  I agree that option 2 (never naturally inferring Literal, even in arguments) would be super annoying and so is probably not good.  Deploying Literals internally, allowing it in type variables (which is helpful for some tensor typing work we're doing) didn't have much backwards compatibility issues (but there were some), which seems to suggest that saying that this is best-effort is the best road to take.
_______________________________________________
Typing-sig mailing list -- typing-sig@python.org
To unsubscribe send an email to typing-sig-leave@python.org
https://mail.python.org/mailman3/lists/typing-sig.python.org/


--
--Guido van Rossum (python.org/~guido)
Pronouns: he/him/his (why is my pronoun here?)
_______________________________________________
Typing-sig mailing list -- typing-sig@python.org
To unsubscribe send an email to typing-sig-leave@python.org
https://mail.python.org/mailman3/lists/typing-sig.python.org/