MFC floating point precision problem

Printable View