Vadim Belman

@rushmorem It looks like you haven't succeed with enums.

@FCO could you, please, recheck with the latest `HEAD`? Though I think what you have here is about roles not closing over their contexts. Because this is intended behavior this...

A quick check shows that the first call comes from optimization and is a result of some kind of typechecking it does. The second is the actual puning. I don't...

> Circling back to the original question: it looks like indeed only roles with positional parameters are the ones that are cached: anything else will just not be cached: https://github.com/rakudo/rakudo/blob/master/src/Perl6/Metamodel/ParametricRoleGroupHOW.nqp#L59...

Unfortunately, I don't see how a typeenv could be re-used. It has to be saved in an attribute of `Metamodel::CurriedRoleHOW`. But `BOOTContext` is not a serializeable thing and being stored...

@coke It's only a single fix. Role parameterization is still done the old style and thus the original issue remains unresolved.

I've been there, tried fixing, until @jnthn pointed out at the fact that maintaining the full chain of `Promise` call "stack" will result in huge memory hog due to closures...

Then `start` would need to know if the context where it's invoked doesn't provide own dynamics. And even if this is implemented then it means one more special case is...

I'm afraid, this kind of solution may produce a lot of WAT in a case of recursive production of chained Promises. While visually these grand-grand-...-Promises may be looking like they're...

Whenever I need to pass in a typed array what I usually do is: ```raku foo( my Int @ = [1,2,3] ); ``` It's a bit cumbersome, but considering that...