On Monday, 18 November 2019 at 21:54:53 UTC, Steven Schveighoffer
wrote:
You are expecting floating point to behave as if it is stored
as a decimal number. It's not.
I was actually asking why 'c' and 'c2' functions behave
differently. After mipri's answer I learned that the default
floating point representation for literals in D is 'real', not
'double', as in C++.
Changing the type to real resolves the different behavior I was
observing.
Just out of curiosity: in C++ a similar code also issues 3704 --
but the same value is returned by 'c' and 'c2', hence my doubt.
Using, in C++, 'long double' and suffixing all floating point
literals with 'l' gave me the same results as my "fixed" D code.
But, yes, thanks for clarifying about math.round.
