Some remarks on measures of noncompactness and retractions onto spheres (Q2460840)

Let \(X\) be an infinite-dimensional normed linear space, \(B(X)\) be its unit ball and \(S(X)\) its unit sphere. It is well-known that there is a retraction \(R\) of \(B(X)\) onto \(S(X)\), equivalently, the sphere is contractible or the unit ball does not have the fixed point property for continuous mappings. One interesting question, and in general open, is: what is the best Lipschitz constant for \(R\)? In this interesting paper, the authors consider the following problem: If \(\gamma\) is a measure of noncompactness, what is the best constant \(k\) such that \(R\) is \(k-\gamma-\)nonexpanding? The authors prove that if \(X= C[0,1]\) and \(\gamma\) is the Hausdorff measure of noncompactness, then \(k=1\) and, using the concept of minimal displacement, they give an upper bound in the general case. In the second part, it is proved that this estimation gives a quantitative estimate for a theorem of Birkhoff--Kellog type and also deduce a condensing version of the Nussbaum fixed point theorem.