Hi, thank you

You were absolutely right.. that does work just like you said.

Here's an element of the puzzle that I left out, and this is where I was getting caught.. this entire thing is running through xinetd. And even though the parent was dieing off, the child still was keeping the standard file discriptors open. As a result, my client that was waiting for an EOF on the socket was never getting it. DOH!

So when I took what I learned from here and the example code and then applied it directly into the real code it worked perfectly as soon as I let my child run Proc::Daemon::Init, which closed the file handles for me.

Thank you everybody for helping me figgure this one out!