reproducibility of errors and failed analyses

[mkofler96]

We should come up with a way of making error reproduction easy. The first thing that comes to my mind is to either store all the steps a user does in its current .vtu file, or at least save all the settings that the user did in the .vtu. For full reproducibility, also the geometry needs to be saved in the .vtu file