I work a lot small tolerances and write floating point exponents a lot. I'd like to suggest that literals that contains an 'e' or 'E' is always a float literal, so that one can write '1e-6' instead of '1.0e-6'. Maybe there any reason it dont work this way now?