Weird floating point rounding - Bug or how to control it correctly

[An Pham via Digitalmars-d-learn]

import std.stdio;

    void main()
    {
        float f = 6394763.345f;
        
        import std.format : sformat;

        char[80] vBuffer = void;
        writeln("6394763.345 = ", sformat(vBuffer[], "%.4f", f));
        
    }

Output
6394763.345 = 6394763.5000