In the prospect of elucidating the physics accessible to an electron-positron linear collider in the range 0.09 to 1. TeV it is of prime importance to be very efficient and not miss any signal. This leads to the question of collecting every interesting event and extracting as much information as possible from every collected event. To achieve that, and considering that most of the physics of interest implies bosons like W, Z or possibly Higgses, the identification of these bosons when they decay into hadron jets is essential. This is the motivation for developing a method which best reconstructs the direction, energy and mass of these di-jets. We believe to have such a method with the so-called "energy flow" or "particle flow" approaches.