Trevisan et al. present that, in many circumstances of the constraint satisfaction problems they examine, the gadgets leading to the strongest possible inapproximability outcomes could