Troeps tries to find plausible values for fields whose values are undefined.
To that extent, it takes advantage of the inference machanisms that are defined for the class fields of the classes to which the object is attached and their superclasses. These inference mechanisms are:
The values that are infered that way are displayed in italic, and, since, there can be several inference methods triggered providing identical or differing values, they are prefixed by the name of the viewpoint under which the value is plausible. The methods are triggered in a total order so there is at most one value by viewpoint.
Of course, these values are not assigned to the actual fields since they are just plausible guesses based on the knowledge base and the actual object attachements and field values.