Continuous variation data: 1:1 or 2:2 weak complexes?