Where should the type information be 221
Where should the type information be 222
and I've managed to write another sentence with blurred meaning. :-) Exactly. And the reason a ** was invented to translate to exponentiation is...
Where should the type information be 223
The function was recognized as essential in numerical mathematics, for which FORTRAN was invented. BTW, the original language is all caps, but the later versions are normally written...
Which is how it is supposed to be read.
Please provide an example of that claim.
Not when F.1 has: "An implementation that defines STDCIEC559 shall conform to the specifications in this annex."
I believe F.9#1 is implying that these specifications may not make much sense for non-IEEE-754 implementations.
Yes. There is no requirement that sin(), cos(), exp(), ... honor rounding directions. Since most implementations of those math functions do not have 0.5 ULP accuracy, it makes little sense for them to honor the rounding mode.
Do you know on what grounds they disagree? Do they believe that buttignment is not required to strip off extra precision and-or range? Do they believe that they can use the value being buttigned in place of the value of the object buttigned to (that is, use the RHS in place of a use of the LHS later)? I buttume you have the same problem with casts.
If the side effects (raising floating-point exceptions) are then lost, or if the value depends upon the rounding mode, and if #pragma FENVACCESS ON is in effect, then they have violated the requirements of C99.
I believe we (the C99 committee) missed that detail. --- Fred J. Tydeman Tydeman Consulting +1 (775) 358-9748 Vice-chair of J11 (ANSI "C") Savers sleep well, investors eat well, spenders work forever.