A remark on Wagner's ring of normal numbers (Q1325095)

\textit{G. Wagner} (Preprint) constructed rings of normal numbers for the first time as follows: Let \(a\) be an odd prime, \(b\) an integer with \(b\geq 2\), and \(\{\alpha_ n\}\) and \(\{\beta_ n\}\) be strictly increasing sequences of positive integers satisfying (G1) \(\lim_{n\to\infty} \alpha_ n/ n\beta_{n-1}= \infty\) and (G2) \(\lim_{n\to\infty} \log \beta_ n/\alpha_ n =\infty\). Then all the nonzero elements of the ring generated by \(\prod_{n=1}^ \infty (\varepsilon_ n a^{-\alpha_ n} b^{-\beta_ n})\) \((\varepsilon_ n=\pm1)\) are normal to base \(b\) and nonnormal to base \(ab\). The author weakens the condition to that \(a\) is any integer with \(a\geq 2\) and \(\lim_{n\to\infty} \alpha_ n/ n\beta_{n-1}= \infty\) instead of (G1). The proof is quite different from that of Wagner.