Separate ruby header files with type information? Seriously? What's the rational behind that? Is it just to make clear that the ruby interpreter doesn't really care about the type information and doesn't use it to improve the code's performance?
With all due respect, but IMHO this is too little and much too late.
replies(2):