# nt.number theory – Generalization of \$lim_{n rightarrow infty} prod_{i=1}^{n}frac{2n-1}{2n}\$ for a character \$chi:mathbb{Z}/s mathbb{Z} rightarrow mathbb{C}^*\$

Playing with some infinite products I came up with this problem, that I’m not able to figure it out by myself. Moreover in the internet it doesn’t seem to appear anywhere.
Maybe it is just an easy consequence of properties of characters that I’m not aware of, anyway thank you in advance for any help/answers/suggestions.

All of us know (it is fairly easy to see) that $$lim_{n rightarrow infty} frac{1 cdot 3cdot dots cdot (2n-1)}{2 cdot 4 cdot dots cdot (2n)}=0$$

Now this fact could be reformulated in this fashion: let $$chi:mathbb{Z}/2 mathbb{Z} rightarrow mathbb{C}^*$$
the only non trivial character of $$mathbb{Z}/2 mathbb{Z}$$, then the expression $$lim_{n rightarrow infty}(prod_{n in mathbb{N}}n^{chi(n : text{mod} 2mathbb{Z})})^{-1}=0$$
More generally, we can perform this construction for every $$s in mathbb{N}$$.

Indeed all we have to do is to consider a non trivial character $$chi:mathbb{Z}/s mathbb{Z} rightarrow mathbb{C}^*$$
and consider the limit
$$lim_{n rightarrow infty}(prod_{n in mathbb{N}}n^{chi(n : text{mod} smathbb{Z})})^{-1}$$

Now it is true that:

1. The value of the limit is finite for every $$s$$ and every non trivial character $$chi$$?
2. If so the value of the limit depends only on $$s$$ or also on $$chi$$?
3. The limit is always a real number? (Possibly $$0$$?)