A zero-dimensional \(F\)-space that is not strongly zero-dimensional (Q2670102)

A space is zero-dimensional if it is a \(T_1\)-space and its clopen subsets form a base for the topology. On the other hand, a Tychonoff space \(X\) is strongly zero-dimensional if any two completely separated sets are separated by a clopen set, that is, if \(A\) and \(B\) are such that there is a continuous function \(f:X\to[-1,1]\) with \(f[A]=\{-1\}\) and \(f[B]=\{1\}\), then there is a clopen set \(C\) such that \(A\subseteq C\) and \(C\cap B=\emptyset\). Equivalently, a Tychonoff space \(X\) is strongly zero-dimensional if the Čech-Stone compactification \(\beta X\) is zero-dimensional. There are by now many examples of zero-dimensional spaces that are not strongly zero-dimensional, even metrizable ones. In this paper, the authors present an example of a zero-dimensional \(F\)-space that is not strongly zero-dimensional. There are many characterizations of \(F\)-spaces but the authors take as the definition the one that at first glance seems quite close to strong zero-dimensionality. A Tychonoff space \(X\) is an \(F\)-space if for every continuous function \(f:X\to\mathbb{R}\) there is another continuous function \(k:X\to\mathbb{R}\) with the property that \(f=k\cdot |f|\). The main part of the paper is devoted to the construction of the \(F\)-space. This construction is inspired by an answer to a question on MathOverFlow, which in turn was inspired by Dowker's example \(M\) in [\textit{C. H. Dowker}, Q. J. Math., Oxf. II. Ser. 6, 101--120 (1955; Zbl 0066.41204)]. The authors first consider an ordered continuum \(K\) with a dense subset \(D\) that can be enumerated as \(\left\langle d_{\alpha}:\alpha\in\omega_2\right\rangle\) in such a way that every tail set \(T_{\alpha}=\{d_{\beta}:\beta\geqslant\alpha\}\) is dense in \(K\). Next they create a sequence \(\left\langle K_{\alpha}:\alpha\leqslant\omega_2\right\rangle\) of split intervals over \(K\). In order to get the desired \(F\)-space, they use, for every \(\alpha\leqslant\omega_2\), the Čech-Stone compactification \(\beta(\omega\times K_{\alpha})\) of the product \(\omega\times K_{\alpha}\) and the remainder \((\omega\times K_{\alpha})^*\). Moreover, the authors discuss some variations of the construction; the \(F\)-space can have arbitrary large covering dimension and its Čech-Stone remainder can be an indecomposable continuum. It is also possible to make the \(F\)-space locally compact. Finally, the following questions to be investigated are posed. \textbf{Question 1.} Is there a subspace of \(\mathbb{N}^{*}\) that is not strongly zero-dimensional? \textbf{Question 2.} Is every zero-dimensional \(F\)-space of weight \(\aleph_1\) strongly zero-dimensional? \textbf{Question 3.} Does the equality \(\textrm{dim}\,X = \textrm{ind}\,X = \textrm{Ind}\,X\) hold for every \(F\)-space of weight \(\aleph_1\)?