YES(?,O(n^2)) Problem: half(0()) -> 0() half(s(s(x))) -> s(half(x)) log(s(0())) -> 0() log(s(s(x))) -> s(log(s(half(x)))) Proof: RT Transformation Processor: strict: half(0()) -> 0() half(s(s(x))) -> s(half(x)) log(s(0())) -> 0() log(s(s(x))) -> s(log(s(half(x)))) weak: Matrix Interpretation Processor: dimension: 1 interpretation: [log](x0) = x0 + 18, [s](x0) = x0 + 4, [half](x0) = x0 + 14, [0] = 12 orientation: half(0()) = 26 >= 12 = 0() half(s(s(x))) = x + 22 >= x + 18 = s(half(x)) log(s(0())) = 34 >= 12 = 0() log(s(s(x))) = x + 26 >= x + 40 = s(log(s(half(x)))) problem: strict: log(s(s(x))) -> s(log(s(half(x)))) weak: half(0()) -> 0() half(s(s(x))) -> s(half(x)) log(s(0())) -> 0() Matrix Interpretation Processor: dimension: 2 interpretation: [1 1] [6] [log](x0) = [0 1]x0 + [0], [1 1] [0] [s](x0) = [0 1]x0 + [1], [half](x0) = x0, [14] [0] = [2 ] orientation: [1 3] [9] [1 3] [8] log(s(s(x))) = [0 1]x + [2] >= [0 1]x + [2] = s(log(s(half(x)))) [14] [14] half(0()) = [2 ] >= [2 ] = 0() [1 2] [1] [1 1] [0] half(s(s(x))) = [0 1]x + [2] >= [0 1]x + [1] = s(half(x)) [25] [14] log(s(0())) = [3 ] >= [2 ] = 0() problem: strict: weak: log(s(s(x))) -> s(log(s(half(x)))) half(0()) -> 0() half(s(s(x))) -> s(half(x)) log(s(0())) -> 0() Qed