This is a rather personal account of my collaboration with Peter Kloeden from 1980 to date. Although our decade-long contact resulted in only two joint papers, Peter has tremendously influenced my and many of my colleagues' work on Random Dynamical Systems by providing innumerous valuable suggestions and ideas, and through his many papers and books.
I will first report on our work on the Lyapunov exponents and rotation number of two-dimensional systems perturbed by telegraphic noise. Then I will describe our results on discretizing a Random Dynamical System at a hyperbolic point. However, it turned out that our assumptions for the validity of our almost sure approximation are unnatural to assume - a consequence of a general result proved by Gunter Ochs.