Floating-point to integer conversion

[Fred Tydeman]

Subject:  Floating to integral conversion
 
My understanding of ANSI C 3.2.1.3, page 36, line 2:
  "When a value of floating type is converted to integral type, the
   fractional part is discarded."
means the same thing as truncate toward zero.  So, for these four cases:
 
Integer    Floating
result     value to convert
 3          3.9
 3          3.1
 3          3.0
-3         -3.0
-3         -3.1
-3         -3.9
 
the integer part is kept and the fractional part is ignored.
 
My understanding appears to be backed up by the Rationale in 3.2.1.3 on
page 35:
  "There was strong agreement that floating values should truncate
   toward zero when converted to an integral type, the specification
   adopted in the Standard."
 
Yet, in reading the C++ draft proposal (X3J16/93-0010, Jan 28, 1993),
section 4.4 Floating and Integral, it says:
  "Conversion of a floating value to an integral type truncates; that
   is, the fractional part is discard.  Such conversions are machine
   dependent; for example, the direction of truncation of negative
   numbers varies from machine to machine."
 
It appears that C++ allows truncate towards minus infinity, as in:
 
Integer    Floating
result     value to convert
 3          3.9
 3          3.1
 3          3.0
-3         -3.0
-4         -3.1
-4         -3.9
 
Am I misunderstanding C, or is C++ changing the meaning from C?
 
These are not IBM's views, these are the personal ones of:
Fred Tydeman, IBM, Austin, Texas (512) 838-3322; fax (512) 838-3484
AIX S/6000 Math library architect & IBM's rep to NCEG (X3J11.1)
Internet: tydemanaibmpa.awdpa.ibm.com    uucp: uunet!ibmsupt!tydeman