Hi, I've been having the `start failed' problem and after snooping around I noticed there was no pidfile yet hald was running. After manually stopping and starting hald (/etc/init.d/hal stop or start) then rerunning dselect (to clear up the failed-config status, which failed) I noticed the pidfile contents did not match the pid.
Manually killing hal (# killall hal), rm-ing the pidfile, then rerunning dselect's install function, seems to have fixed things. Curiously, the other two unstable boxes I run didn't have this problem. The only (for the most part) difference between these boxes is that the one with the problem has /usr mounted via NFS... a similar situation has occurred with exim and the problem was traced to start-stop-daemon not properly stopping the process (exim still fails to stop when shutting down, but seems to be working properly during upgrades). I am starting to believe that NFS and start-stop-daemon don't always get along well, but have no evidence that is the case or any theory as to why. HTH - Bruce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]