"SRV-Style" labels and "IDN"

[Harald Alvestrand]

Mark Andrews wrote:
> In message <8BC845943058D844ABFC73D2220D4665081E4704 at nics-mail.sbg.nic.at>, "Al
> exander Mayrhofer" writes:
>   
>> Hi,
>>
>> I'm currently reading through the idnabis-Documents, and didn't find a
>> clear answer to the following issue:
>>
>> 2.3.2.3 of draft-ietf-idnabis-defs-08 defines an IDN as "name that may
>> contain any mixture of NR-LDH-labels, A-labels, or U-labels".  From what
>> i read, that means that an IDN may not contain a "SRV-style" label,
>> starting with an underscore. A few lines later, it is clarified that
>> such a label is not an "Internationalized Label". There's no text about
>> those labels in the definition of "IDN". Therefore, from the definition,
>> the following example:
>>
>>    _sip._tcp.<A-Label>.<tld>
>>
>> Would not be an IDN. My questions now, unless i have missed something:
>>
>> - Is my assumption correct that the example is not an IDN?
>> - If it is not an IDN, what is it? 
>>     
>
> 	This reflects the mis-use of DOMAIN in IDN.  In reality
> 	IDN's should be named IHN's (International HOST Names).
> 	LDH is a HOST NAME restriction, not a DOMAIN NAME restriction.
>
> 	In general you won't ever enter _sip._tcp.<A-Label>.<tld>
> 	in a application.  You will enter <A-Label>.<tld> and
> 	_sip._tcp will be prepended as part of the discovert process.
>
> 	_sip._tcp.<A-Label>.<tld> is not a host name the same way
> 	that _sip._tcp.example.net is not a host name but both are
> 	legal domain names.
>
> 	Mark
>
>   
we lost this fight (for rational use of "host" vs "domain" in the 
original IDN group back around 2000.

in _sip._tcp.<A-label>.<tld>, it's rational to consider <A-label>.<tld> 
an IDN. I don't have a name for the assemblage.

                  Harald