That works for me! Thanks for your answer. Difficult for humans to reason about + difficult to implement seems like a decent reason to leave it out :). In general, do you think covariant types are more trouble than they're worth? I don't think I've ever encountered them before (or if I did, I didn't realize what they were) so I can't say much on the subject.
> Why restrict it to result parameters--shouldn't we also permit the interface > method's input parameters to be assignable to the type method's input > parameters? It is now much harder to understand whether a type implements > an interface. Yep, I'd imagine it would apply to input parameters as well. This would definitely make it harder for us to reason about whether something implements an interface or not. Didn't fully realize that till you wrote it out. Thanks again! Lucas -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it, send an email to golang-nuts+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
