John Nemeth wrote: > It would also happen if the parent doesn't call waitx(2) to reap > it. That's never supposed to happen. The mimedefang-multiplexor source has a SIGCHLD handler that schedules reaping to happen. I have never seen it fail. Regards, David.