mycroft@gnu.ai.mit.edu (Charles M. Hannum) writes:

>> It is, in fact, not a bug.  It's been the documented behaviour of
>> sockaddr_in for a very long time.
>> 
>> This was originally because sockaddrs were always the same size and
>> were compared with bcmp().  This is no longer the case, and arguably
>> the behaviour should be changed so that it doesn't compare sin_zero.
>> However, doing this does not change the fact that the program in
>> question is broken.

"Be liberal in what you accept...."  The program in question is
broken, but a robust kernel should deal anyway.

		Marc