Application-specific similarity metric

[lu-hub77]

Develop a similarity metric which is adapted to our application, i.e. it is not only considering same / different design decisions, but also the content of the implementation and the kind of change.

Possible questions:

• Are the different synthesis steps equally weighted?
• Do the synthesis steps depend on each other (does one change cause another)?
• Are modifications closely related or broadly distributed?

[lu-hub77]

• "Equally existing" should be weighted higher than "Equally not existing"
• "Equal number of hops" should be weighted higher than "Equal routing decision"