1: \begin{abstract}

2: An $XY$ model with random phase shifts as a model for a superconducting

3: glass is studied in two and three dimensions by a zero temperature domain

4: wall renormalization group which allows one to follow the flows of both the

5: coupling constant and the disorder strength with increasing length scale.

6: Weak disorder is found to be marginal in two and probably irrelevant in three

7: dimensions. For strong disorder the flow is towards a non-superconducting

8: gauge glass fixed point in $2d$ and a superconducting glass in $3d$. Our

9: results are in agreement with recent analytic theory and are inconsistent

10: with earlier predictions of a re-entrant transition to a disordered phase at

11: very low temperature and with the loss of superconductivity for any finite

12: amount of disorder.

13: \end{abstract}

14: