This is particularly problematic when calling FD_SET( -1, ... ), but let's
check it in all functions.

This was introduced with the new API and the fact the net_free() now sets the
internal fd to -1 in order to mark it as closed: now using this information.

We now require support for stdint.h from the compiler.

For some reason select() doesn't seem to work.

This is made possible by the new API where net_accept() gets a pointer to
bind_ctx, so it can update it.

close() may be more meaningful, but free() is symmetric with _init(), and more
consistent with all other modules

Provides more flexibility for future changes/extensions.

On Windows, recvfrom() returns an error code if the destination buffer is too
small to hold the next datagram.

We document that either of recv or recv_timeout may be NULL, but for TLS we
always used recv... Thanks Coverity for catching that.
(Not remotely trigerrable: local configuration.)

Also made me notice net_recv_timeout didn't do its job properly.

- avoids dependency on snprintf
- allows using "smtps" instead of "456" if desired

Some toolchains do not have inttypes.h, and we only need stdint.h which is a
subset of it.

A simple execution of tmp/invoke-rename.pl

Apparently it thinks getsockopt()'s should be a char *, while it's a void *
according to POSIX. Casting to void * avoids the warning.