Is there a reason boost::parser::none::fail() is a runtime error, not a compile time error?

[andreasbuhr]

Hi,

when I mess up my parsers, sometimes I run into none::fail. This is a runtime error. Why not a compile time error with static_assert(false) or static_assert(!std::is_same_v<T,T>) to force late evaluation?

Is there a reason I am missing?