Hilbert's error? (Q1622647)

The Poncelet-Steiner theorem says that whatever can be constructed by straightedge and compass together can be constructed by straightedge alone, if one circle and its center are given. In [Math. Ann. 73, 90--94 (1912; JFM 43.0583.02)], \textit{D. Cauer} published an elegant argument given by Hilbert in one of his lectures that shows that the Poncelet-Steiner theorem cannot be improved in the sense that only the circle without its center is used. The author shows that Hilbert underestimated the problem. His argument was elegant but insufficient. We need to precisely define the notion of a geometric construction which turns out to be nontrivial. However, then we can correct Hilbert's proof.