Joe Eli McIlvain

Discussed in the sync call today, and this looks good. Basically this means that method calls with different receiver caps that direct to the same underlying method (i.e. are "subordinate"...

The explicit recover block here is a red herring - this fails in the same way even if that is removed. The array inference logic here is flawed: https://github.com/ponylang/ponyc/blob/46668e275032586af18f2495d5b99a36cf106951/src/libponyc/expr/array.c#L379-L385 Here,...

> The straightforward solution is to look in all declared interfaces and traits for behaviours when deciding whether the literal should be an actor/object/primitive. Yeah, I think that's what we...

> I would consider the existing functionality a bug. > @ponylang/committer, any input? I don't consider it to be a bug because it's working as designed by Carl. We may...

That is, I believe it makes sense to handle all errors from CLI arg parsing in a first phase, before you proceed to assign those values to your program's data...

Thanks for the detailed issue ticket! As you show, there is an edge case of some kind in which the compiler's logic for statically checking the number of bits to...

Yes, I'll un-assign myself, but if anyone wants help understanding the LLVM code gen stuff in this area of the compiler I'd be happy to assist in that way.

Proposed solution from Jason during the sync call: when doing receiver recovery on a `box` call, we should reify `this` as `ref` instead of `box` inside the call.

I think it's roughly around here: https://github.com/ponylang/ponyc/blob/065d70a9b7116195fbfda37f761fdd1ca4231cd7/src/libponyc/expr/call.c#L272-L273

Note that after [adopting George Steed's viewpoint adaptation model](https://github.com/ponylang/rfcs/pull/122), the example above could be solved with a viewpoint-adapted type. For example, you could have the type parameter `A: Any ref`,...