Vint, I don't think I know any longer, if I ever did, what "IDN" means. Alternatives to Unicode were proposed during the IETF IDN WG lifetime, both as a single normative reference, and as a normative reference. Likewise an intermediate tables redefinition of Unicode, mentioned in my last pointless comment. Then there is the possibility of research on the problems of character repitoires and interoperable data exchange -- before engineering some solution(s). Proposed to the IRTF Chair and rejected. Are there operational issues to attempt to make this thread remotely on point for NANOG? Probably not. Its just bits, and whether the bits are all 0x000 or quasi-random distributions between 0x000 and 0x177 is water under somebody else's bridge. The constraint-space is "solve in applications" and not "solve in infrastructure". The question of semantic scope is interesting in theory, which was the point of my note to Tony Li, if not tractible in a particular context. Eric