Quick close-open cycles problematic in presence of latency even with delay

Hello, I tested the implementation of the release delay 2 years ago and believed that it was a good fix. However, upon revisiting it, I am observing better results from placing that delay after the closure of tty->fds[FD_NET].fd (but before the mutex unlock of access_lock).

I am assuming that this is the actual point where the socket is released and is therefor the correct position for a delay to allow the TCP/network the required time to get ready for the next open event?

Would you consider moving it to this location?

Also, this tool is really excellent btw! Your work on it is greatly appreciated!