Claire Neveu

Also seeing this with the latest version of make-dedicated-lockfile.

I understood @greenrd as saying that `head` should be safe on the `::` **type**. Basically, he takes issue with the fact that the following does not work: ``` scala scala>...

The `@tailrec` annotation produces an error if the annotated function is _not_ tail recursive. If it is tail recursive then the annotation makes no difference. The annotation is basically a...

That would make it impossible to define recursive functions that are not tail-recursive. So you effectively want a wart that forbids recursive functions unless they are tail-recursive, correct?

Yes, I was just looking at that recently. I have a quite old POC that uses scalameta 1.0 to implement semantic-only lint rules in Wartremover that I'm going to go...

I like those exceptions (otherwise the `s` interpolator is basically useless) but I would also include `Boolean`.

I mean, ideally `s` would take some type-class like `Show` but it doesn't and we should be pragramatic here. The representations of most primitives tend to be common among many...

I'm sympathetic to this view but I'd like to offer a counterpoint: ``` scala scala> Json.toJson(List.empty) :15: error: diverging implicit expansion for type play.api.libs.json.Writes[List[Nothing]] starting with method enumNameWrites in trait...

@sellout We have it enabled in almost all of our projects and yes, we use `List`.

The projects are not open source but it's not difficult to comprehend, we just use `List.empty[Warning]` instead of `Nil`, etc.