Sets of ideal convergence of sequences of quasi-continuous functions (Q472303)

This review is based on and reproduced from the introduction of the paper. Let \(\overrightarrow{f}=(f_{n})_{n}\) be a sequence of real-valued continuous functions defined on a metric space \(X\). It is not difficult to show that the set \(C(\overrightarrow{f})\) of all \(x\in X\) such that \((f_{n}(x))_{n}\) converges is \(F_{\sigma \delta }\). On the other hand, \textit{H. Hahn} [Arch. Math. Phys. (3) 28, 34--45 (1919; JFM 47.0238.02)] and \textit{W. Sierpiński} [Fundam. Math. 2, 41--49 (1921; JFM 48.0274.02)] proved independently that for every \(F_{\sigma \delta }\) set \(A\subset X\) there is a sequence \(\overrightarrow{f}=(f_{n})_{n}\) of continuous functions for which \(A=C( \overrightarrow{f})\). Further research (see, e.g., [\textit{I. P. Kornfel'd}, Izv. Vyssh. Uchebn. Zaved., Mat. 1963, No. 4(35), 79--88 (1963; Zbl 0138.28102) (in Russian)] and \textit{J. Lipinski} [Sov. Math., Dokl. 2, 1258--1260 (1961); translation from Dokl. Akad. Nauk SSSR 140, 752--754 (1961; Zbl 0106.26902)], [Fundam. Math. 51, 35--43 (1962; Zbl 0106.26903)]) involved also sets of points where the sequence is divergent to infinity and the like. The full description of these sets was given by \textit{M. A. Lunina} [Mat. Zametki 17, 205--217 (1975; Zbl 0326.40003) (in Russian)]. In the present paper, mainly the result of Lunina [loc.\,cit.]\ is generalized in two directions by the authors. Firstly, the class \(QC(X)\) of all quasi-continuous functions is considered instead of the class of all continuous functions (see Theorem~3). The obtained result is a generalization of Wesołowska's theorem [the second author, J. Appl. Anal. 7, No. 2, 271--283 (2001; Zbl 1002.26004)] on Lipiński's triples for sequences of quasi-continuous functions (see Corollary 4). The second kind of extension of Lunina's theorem [loc.\,cit.]\ consists in substitution of pointwise convergence of \(\overrightarrow{f}\) by ideal convergence with respect to some ideal \(I\) on the natural numbers. \textit{D. Borzestowski} and \textit{I. Recław} proved in [Real Anal. Exch. 35 (2009--2010), No. 2, 479--486 (2010; Zbl 1222.28006)] an ideal version of Lunina's theorem [loc.\,cit.]\ for sequences of continuous functions and for \(F_{\sigma}\) ideals. By using analogous methods, the authors also prove ideal versions of Corollary~4 and Theorem~3.