Weakly open sets in the unit ball of the projective tensor product of Banach spaces (Q2275507)

Let \(X\) and \(Y\) denote nontrivial Banach spaces. \(X\) has the diameter 2 property (or is a diameter 2 space) if every nonempty relatively weakly open subset of \(B_X\) has diameter 2. The situation is so for uniform algebras, for spaces with the Daugavet property and for spaces with infinite-dimensional centralizer), but these three classes are far from covering all diameter 2 spaces. It is natural to ask how the diameter 2 property transfers to structures as subspaces, quotients, sums or tensor products. The paper under review gives strong, positive results of the type ``\(X\widehat{\otimes}_\pi Y\) has the diameter 2 property when both \(X\) and \(Y\) are particular diameter 2 spaces''. The main results are: (1) If \(X\) and \(Y\) both have infinite-dimensional centralizer, then \(X\widehat{\otimes}_\pi Y\) is a diameter 2 space. (2) The assumptions in (1) can be weakened to read ``\(X^{(\infty)}\) has infinite-dimensional centralizer and \(S_{Y^{\ast}}\) contains an element of numerical index one''. (3) For an infinite compact Hausdorff space \(K\), \(C(K)\widehat{\otimes}_\pi Y\) has the diameter 2 property (the same is known to hold for \(L_1(\mu)\widehat{\otimes}_\pi Y\), \(\mu\) atomless, but this time simply because the Daugavet property is preserved). (4) \(X\widehat{\otimes}_\pi Y\) has the diameter 2 property if \(\sup_n\{\dim Z(X^{(2n)})\}=\infty\); here, \(Z(X^{(2n)})\) denotes the centralizer of the \(2n\)-dual of \(X\).