New criteria of some bounded approximation properties (Q652094)

For Banach spaces \(X\) and \(Y\), let the Banach space of all bounded linear operators from \(X\) to \(Y\) be denoted by \({\mathcal L}(X,Y)\). Let \({\mathcal F}(X,Y)\) and \({\mathcal K}(X,Y)\) denote its subspaces of finite-rank and compact operators. A Banach space \(X\) is said to have the approximation property (AP) if the identity map \(I_X\) on \(X\) belongs to the closure of \({\mathcal F}(X,X)\) in the topology \(\tau\) of uniform convergence on compact subsets of \(X\). If \(I_X\) belongs to the \(\tau\)-closure of the subset of \({\mathcal F}(X,X)\) of the operators of norm \( \leq \lambda\), then \(X\) is said to have the \(\lambda\)-BAP. If \({\mathcal F}(X,X)\) is replaced by \({\mathcal K}(X,X)\), then the corresponding approximation properties are called the compact AP and the the compact \(\lambda\)-BAP, respectively. These are well-studied classical approximation properties. Now, if \({\mathcal F}(X,X)\) is replaced by an arbitrary linear subspace \({\mathcal A}(X,X)\) of \({\mathcal L}(X,X)\), then one obtains general versions of approximation properties, which have been studied since the 1980s by O. I. Reinov, N. Grønbæk, G. A. Willis, H.-O. Tylli, and others, including the reviewer. The author rediscovers these approximation properties (unfortunately, this has gone unnoticed by the referee) and names them the \({\mathcal A}\)-AP and the \(\lambda\)-\({\mathcal A}\)-BAP, respectively. From the main results under review, only Theorems 1.1 and 1.2 are essentially new. They state that \(X\) has the \(\lambda\)-\({\mathcal A}\)-BAP (respectively, the \({\mathcal A}\)-AP) if and only if, for every (separable reflexive) Banach space \(Y\), every \(T \in {\mathcal K}(X,Y)\) belongs to the \(\tau\)-closure of \(\{TS : S \in \mathcal A(X,X), \;\|S\| \leq \lambda\}\) (respectively, of \(\{TS : S \in \mathcal A(X,X)\}\)). The other results are known. For instance, Theorem 1.7 of the paper is contained in Corollary 1.4 of [\textit{O. I. Rejnov}, Math. Notes 33, 427--434 (1983); translation from Mat. Zametki 33, No. 6, 833--846 (1983; Zbl 0543.46009)] and Theorem 1.8 of the paper is essentially the same as Lemma 1.2 of [loc. cit.], modulo a standard perturbation argument; Theorems 1.3 and 1.4 of the paper are immediate from [\textit{A. Lissitsin, K. Mikkor} and \textit{E. Oja}, Ill. J. Math. 52, No. 2, 563--582 (2008; Zbl 1185.46009)].