Roger Balian. I would like to present a small part of a study we started around ten years ago with two colleagues—one being in Armenia and the other in the Netherlands and both travelling frequently, this has slowed down our work somewhat. This work is now finished and is published in detail in Physics Reports. Our goal consisted in trying to understand how ideal measurements behaved without going beyond conventional quantum mechanics. In order to specify in which framework we placed ourselves, I will start by recalling the statistical interpretation that I consider to be the minimal interpretation of quantum mechanics, in which one places the least number of things possible. I will then proceed to the presentation of the model and its resolution, with the problems it raises.