Your first point is important, and now that you mention it, I do remember similar names from the Carbon introduction and this paper. I should probably look to D

Your first point is important, and now that you mention it, I do remember similar names from the Carbon introduction and this paper. I should probably look to D&E for a less presumptive set of goals. As for constexpr - you only get ambiguity if you assign a constexpr expression/function call to a non-constexpr value. If you use constexpr variables to hold the results of your computation, compile-time evaluation is guaranteed. I don't think it's that confusing.