Non universality of structure functions and measurement of the strange sea density

We show that there is no real conflict between the two determinations of the strange sea density from the opposite–sign dimuon production and from the difference of the $F_2$ structure functions measured in neutrino and muon deep inelastic scattering. Once non universal sea parton densities are introduced, which take into account the effects of different mass thresholds and different longitudinal contributions, the discrepancy is shown to disappear and both sets of data are simultaneously well reproduced. No need for a large strange sea content of the nucleon emerges.