Text this: Robustness-Related Issues in Speaker Recognition