Malcolm Ryan wrote: > What would be the best way to represent the constraint that {S1, ..., > Sk} is a partition of S? We call it "disjoint union", so it would be rel(space, SOT_DUNION, s1_to_sk, s); http://www.gecode.org/gecodej/doc/ group__TaskIntSetRelOp.html#g9f24705a5a8f13216bd98b10765b26a4 Guido