In this talk, I will explain why the full group of the action of the rationals on the real line is an example of a connected topologically simple Polish group with ample generics, providing a natural answer to a question of A. Kechris and C. Rosendal. If time permits, I will discuss the following open question: is there a Polish group with ample generics which is not quasi non-archimedean ? The talk is based on a joint work with Adriane Kaïchouh.