RE: [idn] The Business Card problem (was: Re: An experiment with UTF-8 domain names)

["Brian W. Spolarich" <briansp@walid.com>]

| In an IDN context, there would be considerable (consideraby more) 
| incentive to try to reverse-map to both "English" and "national" 
| names. Note, for example, that, for routers and similar resources, 
| IA4 names are required by an ITU recommendation that many ISPs feel
| obligated to honor and that integrity of reverse mappings for
| these devices is often considered quite important.

  This is a good point, and perhaps a justification for something like
IPTR (which I've initially thought of as perhaps something that was 
more of an interest to CJK speakers).  ISPs definitely rely on the
IPV4 address to name mappings to determine information about the 
network, including device identity and function and administrative
domain.  Usually there's a geographical and functional heirarchy expressed
by the name components (I know this was the case on the ANS and UUNET
backbones).  This is probably another example of functional 'overloading'
in the DNS, and perhaps a better way to do this is via SNMP.

  This would imply that as a consequence of IDN-related activities
the device management-oriented WGs may need to consider defining MIB
extensions to assist network operators in discovering needed information
about the network.  Easier said than done, I'm sure.

  -bws