A question of divisibility of an analytic germ by a polynomial (Q5951661)

The author answers the following question: if \(f\) is a germ of an analytic function regular in two distingushed variables \(z\) and \(u\) in Weierstrass sense, is there a polynomial in \(z\) and \(u\) with analytic coefficients that divides \(f\)? The answer is no and the construction of the counterexample is based on a generalization of the theorem of Tarski-Seidenberg. However the formulation of the main theorem attributed to Lojasiewicz (Theorem 2.1 in the paper) contains some misprints. The correct statement of Theorem 2.1 follows: Let \(M\) and \(N\) be two affine spaces, the set \(A\) is contained and semianalytic in \(M \times N\), moreover \(A\) is \(S\)-relatively compact and \(N\)-locally semialgebraic, where the space \(N\) is itself a Cartesian product \(R \times S\) and let \(\pi\) be the projection on \(M \times R\) parallel to \(S\), then the set \(\pi(A)\) is \(R\)-locally semialgebraic.