This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: Change definition of complex::norm


nbecker@fred.net writes:

[...]

| Just because the standard leaves it up to the implementation doesn't
| mean we are obliged to take the narrowest interpretation.  I see no
| harm in defining norm(z) = sqr (real(z)) + sqr (imag (z)).

We're not also obliged to choose an implementation which won't
deliver an accetable answer for limit cases.

-- Gaby


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]