The goal of this Comment is to propose a strategic approach to the design of ferromagnetically coupled dinuclear complexes. The theoretical aspect of the problem is discussed at two levels of approximation. At the first level only the metallic unpaired electrons are considered as active electrons; at the second the eventual spin polarization effects from the highest occupied orbitals of the bridging ligands are included. Three situations leading to ferromagnetic interaction are then presented, namely, the strict orthogonality of the magnetic orbitals, the accidental orthogonality and spin polarization effects. For each situation several examples are described. It is emphasized that the accidental orthogonality can be distroyed by a very small change of structural parameter, whereas the two other situations are less sensitive to such modifications. Finally, it is pointed out that in the case of ferromagnetic interaction there is great uncertainty in the energy gaps between low lying states deduced from the magnetic data.