Twentyone years after the proposal of a role of quantum mechanics statistics for the parton distri- butios and some years after the assumption that the boundary conditions for the DGLAP equations imply Fermi-Dirac functios for the quarks and a Planck function for the gluons, it is useful to recall the facts, which con rm the validity of the parametrization implied by those statements. In fact both at small and high 𝑥 the uncertainties in the regions, where the experimental information is scarce, are smaller than the ones obtained by the standard polynomial parametrizations.