> We (VladD2, me and others) discussed this a lot at RSDN, so, we > urgently ask you to issue an error if an implicitly introduced > variable name hides a type. > It will prevent many mistakes, and we do not see any strong reasons > against it. +1