[syslog-ng] Lost tcp connections not reopened

Hi, this is a problem with syslog-ng.1.3.12 on Solaris.... System 1 forwards using tcp to System 2, both using syslog-ng. System 2 syslog-ng is hupped (kill -HUP `cat syslog-ng.pid`) and reinitializes as normal. System 1 loses tcp connection and writes internal message "Connection broken, reopening in 60 seconds". BUT the connection is not reopened and the forwarded data is lost until System 1 syslog-ng is hupped. Then everything is OK again. It would be really nice if the connection were reopened automatically. Thanks, Gary Carroll