Bob Nystrom

> Adding to that, I'm surprised `avoid_dynamic_calls` doesn't trigger on the `message.$1` within the `case` block by the way. There's no dynamic call when destructuring the fields in the pattern...

> I _think_ named constructors predate metadata I believe you're correct.

> This is basically a request for "expressions statements" #132, which allows statements inside an expression, which still evaluates to a value. Note that this isn't *entirely* a request for...

> This particular situation has come up _a lot_ for me when converting Sass to Dart 3 style. I'll have a nice lovely switch expression going, and then I'll realize...

> I really don't like using `=>` with `void`, and fitting one one line is the _only_ reason to use `=>` for that!) For what it's worth, my impression is...

I agree that this is a problem, but I think it's somewhat fundamental and inescapable. > When a user defines a field in one place, they likely do not realise...

> I find it very unappealing that it is possible to call a method that is not implemented in a static language without a compile time error (I understand why...

Shadowing makes a lot of sense in Rust because the language also has ownership semantics, basically a form of linear typing. In Rust, you can have functions that *consume* a...

I could be wrong, but I don't think this is worth doing. Using `_` is quite a bit shorter than `default`. Honestly, I wouldn't be surprised if `_` became the...

> They have different semantics(one _can_ and should be allowed to write `default` case at top but `_` always comes last). There is no value in having a `default` case...