abstract: At a meeting in Pisa two years ago, I described a counterexample to the relative Manin-Mumford conjecture, due to the existence of Ribet sections on certain semi-abelian schemes (and immediately turned by B. Edixhoven into a geometric statement). In the present talk, I will report on a joint work with D. Masser, A. Pillay and U. Zannier, which shows that these sections are the only obstruction to the validity of the conjecture for semi-abelian surface schemes.