I can set a "X,Y tolerance" or use default when I create a new feature class.
I can set a "X,Y tolerance" or use default when I create a new feature dataset.
I can set a "Cluster tolerance" or use default when I create a new topology within a dataset.
What are the differences, and what is affected if there are different settings (for example, I have a topology with a .0034 cluster tolerance in a dataset set to .00328 X,Y tolerance with feature classes of .00328 X,Y Tolerance settings).