[ietf-dkim] Internationalized domain names
Paul Hoffman
phoffman at proper.com
Wed Jul 19 16:20:28 PDT 2006
At 4:08 PM -0700 7/19/06, Douglas Otis wrote:
>On Jul 19, 2006, at 10:29 AM, Paul Hoffman wrote:
>
>>>
>>>Agreed.
>>>
>>>Calling this an ACE label which is identified by the "xn--" prefix
>>>may avoid some of the confusion.
>>>
>>>This could also be written as:
>>>
>>> Internationalized domain names MUST be converted using the
>>> steps in section 4 of [RFC3490] into "ACE" labels using the
>>> ToASCII function.
>>
>>No, that's wrong: only the labels that have non-LDH characters are
>>converted. Section 4 in IDNA covers this correctly.
>
>Good point, but not exactly right either.
Please show *exactly* where the following is "not exactly right":
Internationalized domain names MUST be converted using
the steps in section 4 of [RFC3490] using the ToASCII
function.
>A term for international domain names encoded into ACE labels would
>provide a useful reference, rather than calling this puny-code.
The proposed text does not use the term "puny-code".
>The '_' underscore character is also not one of the 37
>case-insensitive LDH (letter, digit, or hyphen) characters either,
>but the intent is not to include this label within this requirement.
Correct, and irrelevant.
> While domain name encoding is specifically mentioned with the i=
>parameter where there are essentially no character restrictions for
>subdomains, this statement also applies to the d= tag as well.
Correct, and irrelevant. You cannot use "_" in labels in domain names
in either i= or d=.
>Perhaps:
>
> Internationalized domain names MUST be converted using the
> steps in section 4 of [RFC3490] into "ACE" labels when
> appropriate by applying the ToASCII function for "stored stings."
> This requirement does not apply to the "_domainkey" label.
No. Changing a simple, correct statement into a confusing one with
new terminology is a bad practice. Please stop.
More information about the ietf-dkim
mailing list