[en] After the successful basic solution of the Nb₃Ge and Nb₃Si problems, optimization strategies and quality criteria were developed for high Tc superconducting (HTCS) samples in analogy. After improvements of the sputtering system, the preparation parameters were tightly controlled, leading to samples with a reproducible linear slope of the specific resistance ρ(T). Further improvements by us and Poppe et al. lead to the preparation of 123 films with non-linear ρ(T) curves above 120K. A generalization to a T² behavior of ρ(T) leads to the further generalization that defect free HTCS are essentially two dimensional in nature, with all ensuing consequences. The momentum transport of the moving charge cariers requires, because of momentum conservation, a mechanism to transfer momentum to the lattice (momentum relaxation). The qualitative theory developed for two dimensional (2D) systems shows that the electron-phonon interaction is strongly suppressed and hence that the weaker electron-electron interaction dominates. That theory makes numerous predictions, which so far are confirmed by experiments. In particular, it accounts for the extreme temperature dependence of the mean free path and the Hall effect. It is inferred that the mechanism of HTCS is predominantly due to electron-electron interaction - probably of bipolaron nature, as indicated by experiment. At this point experiment meets just a few theories, that conform with the experimental facts and above theory, all based on the quasi-particle theory of electron-electron interaction. The theory of Newns, Pattnaik and Tsuei should have a T² shape of ρ(T) and, because E_F lies near a nearly logarithmic van Hove singularity, accounts for the high Tc. Ruvalds and Virosztek's theory predicts the T² shape of ρ(T). (orig./GSCH)