...or when SIGTERMed. The former coredump doesn't happen every time; the latter coredump seems to occur consistently. The core file generally ends up in either / or /var/lock. papd from 2.1.3, running on the same system and configured the same way, doesn't have this problem. Red Hat 6.0-Intel, kernel 2.2.12; netatalk compiled from source (both the 2.1.3 and the pre-2.1.4 versions) with essentially the same Makefile configs. Since this is a production system, I've gone back to 2.1.3. But... I'm curious, since I haven't seen any other mentions of this one on the list: anyone else seen this? --Benjamin Gilbert
