It looks to me that this release does not honour the flush_timeout setting. I have flush_timeout(1000) on my syslog client and my syslog server, but every time I log something on the client, it appears nearly instantly in my "tail -f" output on my server. Can anyone else confirm or refute this? Evan.