Frames and Their Names

 

Actually, this is not quite right. For initial presentation, we have allowed a certain amount of confusion between the thing itself, and its name. Really, the frame for Adam would look something like:

 
frame-27:
   name:    ("Adam")
   sex:     (frame-2)
   spouse:  (frame-28)
   child:   (frame-29 frame-30 frame-31)

A frame has two different types of name.

• Every frame has a single true name (tname), which is a symbol referring uniquely to that frame. In the Lisp implementation, the tname is the Lisp symbol on whose property list the frame structure is stored. In the example above, it is frame-27.
• A frame can have any number of public names (pnames), which are strings, and need not be uniquely referring. Public names are stored as values in the name slot of the frame. There is also an indexing mechanism that allows frames to be retrieved, given a public name.

True names are the pointers from one frame to another that actually represent the structure of the knowledge base. Public names are for communication with other agents.

For ease of debugging and interaction, when a frame is created with a simple public name, Algernon attempts to generate a true name that has the same or similar printed representation. Similarly, the user interface attempts to retrieve the desired frame, whether you type its true name or its public name. However, this should be treated as a happy coincidence when it happens.