Let $(X,0)$ be a normal surface singularity. An let $\pi: \tilde{X} \to X$ be the minimal resolution. Now, we can apply a result of Oliveira (exploiting previous work by Laufer) and obtain a 1-parameter deformation $d: Y \to D$ where the central fiber is isomorphic to $\tilde{X}$ and $D$ is some small disk.
On this space we apply Remmerts reduction and obtain a normal space $\tilde{Y}$ on which (because it is normal) the map $d$ induces a map $s:\tilde{Y} \to D$ which is actually a smoothing of the central fiber. This central fiber has certainly $(X,0)$ as its normalization but it cannot be normal (since we would have proven that all normal surface singularities admit a smoothing).
My question is, if the Remmerts reduction applied on $(Y,0)$ only modifies the space within the central fiber, how is it possible that , restricted to the central fiber, it is different from the resolution map which only contracts the exceptional set of the central fiber as well?
