Charles mentioned in IRC that Groovy has multiple "self" notions
depending on if you want to refer to the closure or the surrounding
object. Personally I don't see much use in accessing fields and self
of the closure object itself, but maybe there's a use case for that
which I don't see. It might be nice to try to emulate Ruby's approach
and shield programmers from the fact that closures are separate
classes, but (a) how hard would that be to implement? and (b) can that
be done seamlessly, or will it be a leaky abstraction? If it can't be
done seamlessly with "self" and "@var" then it may be necessary to
introduce new ways to reach into the surrounding object.
Thoughts?
-Phil