Before trying to add a network connection, the system may present to the user a list of local devices to select from. So that entries are not presented in this list that the NP would regard as invalid, the system will call NPValidLocalDevice first. If all of the valid devices for the network are described by the registry entry, then the NP does not need to support this routine. However, if the network has its own dynamic restrictions upon valid device ranges, then it should support NPValidLocalDevice to indicate this range to the system. If a device is currently in use, and the NP could form a connection to it were it not, the NP should return that the device is valid to ensure correct error reporting.