YES(?,O(n^3)) Problem: b(c(x1)) -> c(b(x1)) c(b(x1)) -> a(a(a(x1))) a(a(a(a(x1)))) -> b(c(x1)) Proof: RT Transformation Processor: strict: b(c(x1)) -> c(b(x1)) c(b(x1)) -> a(a(a(x1))) a(a(a(a(x1)))) -> b(c(x1)) weak: Matrix Interpretation Processor: dimension: 1 interpretation: [a](x0) = x0 + 6, [b](x0) = x0, [c](x0) = x0 + 3 orientation: b(c(x1)) = x1 + 3 >= x1 + 3 = c(b(x1)) c(b(x1)) = x1 + 3 >= x1 + 18 = a(a(a(x1))) a(a(a(a(x1)))) = x1 + 24 >= x1 + 3 = b(c(x1)) problem: strict: b(c(x1)) -> c(b(x1)) c(b(x1)) -> a(a(a(x1))) weak: a(a(a(a(x1)))) -> b(c(x1)) Matrix Interpretation Processor: dimension: 1 interpretation: [a](x0) = x0 + 2, [b](x0) = x0 + 8, [c](x0) = x0 orientation: b(c(x1)) = x1 + 8 >= x1 + 8 = c(b(x1)) c(b(x1)) = x1 + 8 >= x1 + 6 = a(a(a(x1))) a(a(a(a(x1)))) = x1 + 8 >= x1 + 8 = b(c(x1)) problem: strict: b(c(x1)) -> c(b(x1)) weak: c(b(x1)) -> a(a(a(x1))) a(a(a(a(x1)))) -> b(c(x1)) Matrix Interpretation Processor: dimension: 3 interpretation: [1 1 0] [0] [a](x0) = [0 1 1]x0 + [0] [0 0 0] [4], [1 3 3] [0] [b](x0) = [0 1 1]x0 + [6] [0 0 0] [4], [6] [c](x0) = x0 + [2] [0] orientation: [1 3 3] [12] [1 3 3] [6] b(c(x1)) = [0 1 1]x1 + [8 ] >= [0 1 1]x1 + [8] = c(b(x1)) [0 0 0] [4 ] [0 0 0] [4] [1 3 3] [6] [1 3 2] [4] c(b(x1)) = [0 1 1]x1 + [8] >= [0 1 1]x1 + [8] = a(a(a(x1))) [0 0 0] [4] [0 0 0] [4] [1 4 3] [12] [1 3 3] [12] a(a(a(a(x1)))) = [0 1 1]x1 + [12] >= [0 1 1]x1 + [8 ] = b(c(x1)) [0 0 0] [4 ] [0 0 0] [4 ] problem: strict: weak: b(c(x1)) -> c(b(x1)) c(b(x1)) -> a(a(a(x1))) a(a(a(a(x1)))) -> b(c(x1)) Qed