We argue that Kantor and Kardar's assertion that their simulation results contradict our criterion for the localization of a softly constrainted ideal polymer is incorrect. Our criterion is inapplicable to the model used in these simulations. We argue that our criterion for localization of as polymer with internal constraints strongly suggests that the models with soft and hard constraints belong to different universality classes. The size measure used by Kantor and Kardar is also not adequate for studying polymer localization. We also show that, in spite of these contrasts, combining the simulation results with our criterion for ideal chains seems to uncover some intriguing physics.