One way it's different is that there is rich support for specifying what should happen when subsystems crash. You make what's called a "supervision tree" where supervisors are notified when a child process crashes, and can choose what to do -- restart that child, restart all similar children, halt execution, do something completely custom, etc.

Scala's Akka library is another place you can find this (although a bit harder to configure) but I don't know many other languages that offer this.