Any updates on this? Updated PR with respect to master

I'm not sure adding this note is actually helpful. The root problem seems to be that the current error is just confusing -- it says (for example) argument has incompatible type "**Dict[str, float]"; expected "int" where it should explain that the dict's value type is not compatible with all of the function's argument type, or something like that. It is easy to come up with examples where the note is just wrong, e.g.

def f(a: int): ...
f(**dict(a=3.14))

The note is needed in cases where the Dict is untyped and inferred type becomes **Dict[str, object]. In such cases, the dict needs to explicitly typed. e.g.

def f(a: int, b: str): ...
f(**dict(a=1, b="some_str"))

Even though types are correct, it fails.

In cases where inferred dict type is not **Dict[str, object], we can modify the error message to explain the dict's value type is not compatible with functions arguments. Would that work?

The note is needed in cases where the Dict is untyped and inferred type becomes **Dict[str, object]. In such cases, the dict needs to explicitly typed. e.g.

def f(a: int, b: str): ...
f(**dict(a=1, b="some_str"))

Even though types are correct, it fails.

In this case we should do what Jukka suggested in his first bullet in #8874: "Generate a single error message instead of multiple ones."

In cases where inferred dict type is not **Dict[str, object], we can modify the error message to explain the dict's value type is not compatible with functions arguments. Would that work?

I'm thinking there are several cases.

• The key type is not compatible with str -- that should be a separate error message
• The value type is not compatible with one or more argument types -- that should be another error message

In the example you give, maybe the error should be

error: first **kwargs argument to "f" has value type "object", incompatible with formal parameter 1 ("int") and 2 ("str")

(However, I don't know if all the information needed to format it that way is actually available.)

Hi, any progress on this PR?

Closing due to inactivity.