\(p\)-adic logarithmic forms and a problem of Erdős (Q392797)

This is a continuation of a series of six papers of the present author on linear forms in \(p\)-adic logarithms which had appeared between 1989 and 2007 (see the review of the last paper in the series so far [\textit{K. Yu}, Forum Math. 19, No. 2, 187--280 (2007; Zbl 1132.11038)]). According to the author, he re-examined this sixth one `thoroughly and achieved..., through very detailed work,' the following three refinements. 1) The use of the Vahlen-Capelli theorem, as in the six preceding papers (apart from the first) has been removed from the \(p\)-adic theory of logarithmic forms (of length \(\ell\), say). This implies that a quadratic extension of the ground field can be avoided (in case \(p>2\)) leading to a gain of the factor \(2^\ell\) in applications. For example, C. L. Stewart made essential use of this refinement in his work [\textit{C. L. Stewart}, Acta Math. 211, No. 2, 291--314 (2013; Zbl 1362.11070)] on divisors of Lucas and Lehmer numbers. 2) In the estimates for \(p\)-adic logarithmic forms, the dependence on the parameter \(p\) is now improved. Namely, a reduction from \(p^2\) to \(p\) was achieved in case \([\mathbb{Q}(\alpha/\beta):\mathbb{Q}]=2\) and \(p\;(>\!2)\) is inert in \(\mathbb{Q}(\alpha/\beta)\) (compare the preceding review on Stewart's paper). 3) Finally, important improvements on the numerical constants appearing in several results could be made.