>
> That's the solution you've picked.
>
I still haven't figured out how to actually do this in the configuration
file.
> What is the reasoning which lead you to pick this solution?
>
I am going to be using syslog to log messages from an app I am writing.
I need to be able to
have the messages start going into a new file at an arbitrary time.
> If nothing else, there are significant security and denial of service
> issues inherent in this idea, especially when accepting UDP syslog
> messages, trivially spoofable.
>
I am only going to be handling messages from my app, not a catch-all
source.