@Vor If two planners both running the same deco algorithm with the same Gf are giving you different results then that is a serious problem, that suggests there is a bug and one of them is wrong. Is it possible the other one is uing *b varient? That is the only way i could see for it to be explained without it being a bug.