My view is as follows: Although in the paper with Boaz a constraint
associated with a strategy edge throws out nodes and edges from the
class graph, it is useful to be able to change the edge names
when a concrete class graph is known.

A strategy: from Company bypassing -> *,retirement,* to Money
should be possible to map to:
from Firma bypassing -> *,pension,* to Geld

-- Karl

>From dougo@ccs.neu.edu Fri Oct 30 15:33:17 1998
>From: Doug Orleans <dougo@ccs.neu.edu>
>Date: Fri, 30 Oct 1998 15:27:17 -0500 (EST)
>To: dem@ccs.neu.edu
>Subject: edge name maps
>
>Should you be allowed to put edge names in name maps also?  This would 
>be useful for adding an edge-name constraint on an strategy edge
>(e.g. "bypassing -> *,foo,*") that would be mapped onto an edge name
>in the class dictionary (e.g. "foo = spouse").  Or is it not likely
>that someone would want to add an edge-name constraint to a strategy
>graph unless he has a particular class graph in mind?
>
>--Doug
>