Real vs Double vs Float
The "real" type is stated to have double precision and 64 bits. Why not call it a double then, would be consistent with IEEE 754.
Proposal (Rich Heidtman 2021-07-05)
USE The name of basic floating point type: float
USE IEEE7854 double precision floating-point.
Lexical Syntax (Rich Heidtman 2021-07-05)
<float> ::= ["-"<non-zero-digit>]<digit>*"."<digit>* | <digit>*"."<digit>*