A parameter has a number of states. Initially, a parameter has no value, and is said to be in the
ground state. When the parameter has been given a value, the parameter moves to the active
state. A parameter acquires a value in the first instance by looking for one supplied on the
command line. Failing that, the parameter system will attempt to get a value, for example by
prompting or adopting a fixed or dynamic default. If an application obtains a value from a
parameter already in the active state, the existing value for that parameter is returned. To
obtain a further value the application must first cancel the parameter and move it to the
cancelled state. When the application next gets a value for that parameter (within the same
invocation), the parameter returns to the active state. When an application completes the
parameter returns to the ground state. PAR also permits users of your application to enter a null
value for a parameter (a ! in the ADAM parameter system). This puts the parameter into the
null state, and can be used to select a special value or control the application in a special
way.