We study a relation recently derived by K. Kondo at zero momentum between the Zwanziger's horizon function, the ghost dressing function and Kugo's functions u and w. We agree with this result as far as bare quantities are considered. However, assuming the validity of the horizon gap equation, we argue that the solution wð0Þ ¼ 0 is not acceptable since it would lead to a vanishing renormalized ghost dressing function. On the contrary, when the cutoff goes to infinity, uð0Þ ! 1, wð0Þ ! À1 such that uð0Þ þ wð0Þ ! À1. Furthermore w and u are not multiplicatively renormalizable. Relaxing the gap equation allows wð0Þ ¼ 0 with uð0Þ ! À1. In both cases the bare ghost dressing function, Fð0; ÃÞ, goes logarithmically to infinity at infinite cutoff. We show that, although the lattice results provide bare results not so different from the Fð0; ÃÞ ¼ 3 solution, this is an accident due to the fact that the lattice cutoffs lie in the range 1-3 GeV À1 . We show that the renormalized ghost dressing function should be finite and nonzero at zero momentum and can be reliably estimated on the lattice up to powers of the lattice spacing; from published data on a 80 4 lattice at ¼ 5:7 we obtain F R ð0; ¼ 1:5 GeVÞ ' 2:2.