I am writing a mud server in perl and I am worried about securety on the machine that will be running this app. It binds to port 23 (telnet) which requires root privilages which makes the potential dammage caused by malicious parties great. I do not trust my programming enough to risk my network.
I have seen non perl applications, such as apache and bind actually change the executing user in mid execution. (Although apache does this by running more processes, I would still like to know if it is an easy viable solution).
The script I am writing is based around POE::Component::Server::TCP to dispatch jobs from multiple clients which may complicate the matter.