Leo White

I've started to skim over some of this now. The first thing I would say is, even though the original modular implicits prototype did it this way too, the type...

I haven't had time to look at the new code yet, but based on your description I would suggest a slightly different approach for the scopes/iterators. Rather than having mutable...

Another good example is probably something like: ```ocaml let f (x : (({M : S} -> M.t * { N : S } -> M.t * 'a) as 'a)) :...

I think that using variables like that in a principal way would require a different notion of type variable, with different scoping and behaviour. It's not enough to change the...

> And try ... finally%lwt ..., etc. To be clear, I wasn't proposing that `try ... finally ...` would be used by the user instead of a ppx -- although...

That seems worth supporting too. Although for my use case I'd like to take a single trace, view it in its entirety at a low resolution, and then view subsections...

I would recommend not writing any PRs proposing changes to the match or if syntax without some indication from upstream that they will be accepted. It is extremely unlikely that...

I think that it is probably better to include this kind of functionality in a library that schedules concurrent tasks across multiple domains (e.g. [eio](https://github.com/ocaml-multicore/eio) (I think)), for two reasons:...

Here's an example based around the same basic idea: ``` # module Foo (X : sig type t end) = struct type 'a t = 'b * X.t constraint 'a...

I do. I have written the patch to make `module type of Foo` produce the type currently produced by `module type of struct include Foo` in the past, and used...