Abstract
The numerical method of Shugard, Weeks, and Gilmer is shown to contain systematic errors that prevent it from testing theories of planar and roughening models.