Non-stratifiability of \(C_{k}(X)\) for a class of separable metrizable \(X\) (Q876545)

In 1961, \textit{Jack Ceder} defined the classes of \(M_i\)-spaces, \(i = 1, 2, 3\) in [Pac. J. Math. 11, 105--125 (1961; Zbl 0103.39101)]. A regular \(T_1\)-space is called \(M_1\) if it has a \(\sigma\)-closure preserving base while a regular \(T_1\)-space is called \(M_2\) if it has a \(\sigma\)-closure preserving quasi-base. Later it turned out that the classes of \(M_2\) and \(M_3\)-spaces were the same and now they are usually known as stratifiable spaces. Clearly an \(M_1\)-space is stratifiable, but the main question about stratifiable spaces remains open: is a stratifiable space an \(M_1\)-space? There have been a number of interesting attempts, though not entirely successful, on this question. In [\textit{T. Mizokami} and \textit{N. Shimane}, Topology Appl. 105(1), 1--13 (2000; Zbl 0954.54011)], it has been proved that a stratifiable \(k\)-space is an \(M_1\)-space; while in [\textit{P. M. Gartside} and \textit{E. A. Reznichenko}, Fundam. Math. 164, 97--114, (2000; Zbl 0971.46012)], it has been proved that if \(X\) is separable and completely metrizable, then \(C_k(X)\) is stratifiable. In particular, \(C_k(\mathbb{P})\) (\(\mathbb{P}\) stands for the space of irrational numbers) is stratifiable. It appears that \(C_k(\mathbb{P})\) may be a prime candidate for a negative solution to the aforementioned problem of whether every stratifiable space is \(M_1\). So in the context of stratifiability of \(C_k(X)\), the author has posed the following problem in the paper under review: Problem 1. Assume \(X\) is a separable and metrizable space and \(C_k(X)\) is stratifiable. Must \(X\) be completely metrizable? Gartside and Reznichenko have conjectured a positive solution to Problem 1, which easily reduces to the 0-dimensional case [\textit{G. Gruenhage}, Metrizable spaces and generalizations. Recent progress in general topology II. Based on the Prague topological symposium, Prague, Czech Republic, August 19-25, 2001. Amsterdam: Elsevier. 201-225 (2002; Zbl 1029.54036), Proposition 27(3)]. In order to have such a positive solution, the author has proved the following interesting result in the paper under review. Theorem 1. Let \(X\) be a 0-dimensional separable metrizable space such that every compact subset of \(X\) is countable. If \(X\) is not scattered, then \(C_k(X)\) is not stratifiable. Since every scattered space is completely metrizable, the only restriction in Theorem 1 in relation to Problem 1 is in the assumption that every compact subset of \(X\) is countable. But Theorem 1 immediately answers negatively the query of Gary Gruenhage made in [loc. cit.]: is \(C_k(\mathbb{Q})\) stratifiable? The paper under review is highly stimulating and a milestone in the study of stratifiable spaces.