Greetings again. Thanks for the early comments on nameprep-01; we have incorporated them into nameprep-02, which was just released. For those of you carefully looking at the characters in Appendixes E and F and wanting to see the character names, please see: <http://www.i-d-n.net/draft/nameprep-appE-input.txt> <http://www.i-d-n.net/draft/nameprep-appF-input.txt> There was a pretty serious bug in Appendix F in nameprep-01: we accidentally included U+002D (hyphen) in the prohibited list. That has been fixed in -02. If anyone is creating test vectors for nameprep, please get in touch with me. I would like to have publicly available lists of names that should pass nameprep, and names that shouldn't. Both lists should many edge cases. For example, the list of names that should pass should have examples of names that pass unmodified, names that pass modified due to each different type of mapping, names that are modified only by mapping, names that are modified in the mapping step and modifed again in the normalization step, and so on. We also need tests for names that include unassigned characters so that people who are creating client software and authoritative name server software can be sure that they get different results. --Paul Hoffman, Director --Internet Mail Consortium
