Re: My schedule, 2.4 schedule

[Tor Lillqvist <tml iki fi>]

Matthias Clasen writes:
 > What would be the point of including the api without a working
 > implementation ?

The trivial implementation would be a working one, in the sense that
it would work as well as current code that manipulates GPollFD::fd
directly does. I.e., just fine on Unix, and rather badly on Windows.

(The trivial implementation would be:

void
g_pollfd_set_fd (GPollFDType type,
 		 gint        fd,
		 GPollFD    *pollfd)
{
  pollfd->fd = fd;
}

void
g_pollfd_unset_fd (GPollFD  *pollfd)
{
  pollfd->fd = -1;
}
)

--tml