the text for Relation class AllIndexes method says
Returns an array of all indices contained in the Relation. If the Relation
contains multiple indices of the same value the duplicates will be eliminated
from the returned array.
The method actually returns all Indexes and duplicates are not eliminated.
Rick says the method behaves correctly, so the Docs must be wrong.
perhaps it would be worth giving an example of how to get all distinct indexes, like this:
do index over .set~new~union(MyRelation~allIndexes)