1: \begin{abstract}

2:   We prove a general Mosco convergence theorem for bounded Euclidean domains satisfying a set of mild geometric hypotheses.

3:   For bounded domains, this notion implies norm-resolvent convergence for the Dirichlet Laplacian which in turn ensures spectral convergence. 

4:   A key element of the proof is the development of a novel, explicit Poincar\'e-type inequality.

5:   These results allow us to construct a universal algorithm capable of computing the eigenvalues of the Dirichlet Laplacian

6:   on a wide class of rough domains. Many domains with fractal boundaries, such as the Koch snowflake and certain filled Julia sets, are included among this class.

7:   Conversely, we construct a counter example showing that there does not exist a universal algorithm of the same type capable

8:   of computing the eigenvalues of the Dirichlet Laplacian on an arbitrary bounded domain.  

9: \end{abstract}

10: