<@U0ANUS2BA> "take the class out of the T's type” ...
# reflect
d
@udalov "take the class out of the T's type” makes sense, thanks. I was thinking though that consistency would then suggests that
List<Something>::class
would be allowed as well, with the same semantics. although I for one preferred it to be an error, and thought it could be indicated at the call site instead of the declaration one (which would rule out valid use cases as you pointed out). maybe I have a too naive view of how inline functions are compiled though and that’s not doable ?