Leaf Petersen

> I think if we want to go down this path we should start with usability testing to ensure that it is a clear win for readability and maintainability, even...

I'm a little nervous about this proposal, because it pushes on an area where we are constrained by the limited set of union/intersection types we support. In general, unions and...

Per @nshahan the current implemented semantics for dart2js, VM and DDC are as follows: ``` class C { Function get call => () { print('getter call function'); }; } main()...

I filed an issue [here](https://github.com/dart-lang/language/issues/3575) to explore a corner of this that I believe is not covered above.

I think if you think about it, this is the only correct semantics (or at least the only consistent semantics). Patterns execute "outside in". Would you expect this program to...

> These rules have the emergent property that when there's a "choice" of when to apply a coercion (e.g. whether to apply a coercion to the target of a cascade...

@lrhn > ```dart > int x = (b::bool ? i::int : d::dynamic) :: int; > ``` I don't know where you're going with this. As written, this is an unsound...

> TL;DR: I agree that what you present matches the intented currently specified behavior. I’m not sure that behavior is _well_ specified, and we already know it’s not what’s implemented....

> I have been under the impression that we generally consider implicit downcast, and implicit coercions in general, to be (intended to be) specified by _something like_: > > >...

Just a note in passing for the record. F-bounded types will, as usual, require some level of approximation here. For example, if we wish to compute the greatest closure of...