Gavin King
Gavin King
So, look, I'm questioning whether it's really a good idea to push this into the language: - if we were to implement #7058, and finish the implementation of #6615 (i.e....
Given the work on #7058 and #6615, I feel like the pressure for this feature is significantly alleviated.
Thanks for reporting.
I can also reproduce this problem with `VariableAnnotation`. The difference between `VariableAnnotation` and `SharedAnnotation` is that `SharedAnnotation` has a union-typed argument to the type parameter `ProgramElement` of `OptionalAnnotation`.
Alright, so the issue here is rooted in the fact that `ProgramElement` is (correctly, it seems to me, according to the rules, given that it occurs only contravariantly in the...
I can also fix the code by introducing an intervening interface which redeclares `ProgramElement` as an *invariant* type parameter: ```ceylon interface MyOptionalAnnotation of Annot satisfies OptionalAnnotation given Annot satisfies MyOptionalAnnotation...
Can't we just automatically promote them to default access?
Package access, I mean.
To be clear: this change described would break some perfectly-reasonable code.
Note: if I don't hear lots of expressions of support for making this change, I'm not going to do it :-)