Phonon or electron mediated weak BCS attraction is enough to have high crit
ical temperature if a van Hove anomaly is at work. This could apply to elec
tron doped compounds and also to compounds with CuO2 planes overdoped in ho
les, where T, decreases with increasing doping. If phonons dominate, it sho
uld lead to an anisotropic but mainly s superconductive gap, as observed re
cently in overdoped LaSrCuO1 and probably also in electron doped compounds.
If electrons dominate, a d-gap should develop as observed in a number of c
ases. In the underdoped range, the observed decrease of T-c with hole dopin
g can be related in all cases to the development of antiferromagnetic fluct
uations which produces a magnetic pseudogap, thus lowering the density of s
tates at the Fermi level. The observed mainly d-superconductive gap then ca
n be due to a prevalent superconductive coupling through antiferromagnetic
fluctuations; it could also possibly be attributed to the same phonon coupl
ing as in the overdoped range, now acting on Bloch functions scattered in t
ile magnetic pseudogap. More systematic studies of superconductive gap anis
otropy and of magnetic fluctuations would be in order.