Re: [idn] Needed changes to IDNA before last call

[Dan Oscarsson <Dan.Oscarsson@trab.se>]

>> The draft: idn-idna:
>> The ToASCII must be applied to ALL labels containing non-ASCII.
>
>Rule 1 already says:
>
>    Whenever a domain name is put into a generic domain name slot, every
>    label MUST contain only ASCII characters.
>
>"every label".

Yes, but not all labels in DNS are "generic domain name slot" from what
I can understand. For example the e-mail name label in the SOA record.

>
>> The steps need a slight change:
>>  - character restrictions will apply depending on label, it could be
>>    host name or e-mail name.
>
>Step 3 of ToASCII already says "If the label is part of a host name (or
>is subject to host name syntax rules) then perform these checks".
>
>When you say "e-mail name" do you mean the local part (before the @)?
>That's outside our jurisdiction.  Another working group will have to
>decide how to internationalize email local parts.  We can only tell how
>to deal with domain names.

That is right, but the ACE encoding is not what defines the way to
internationalise host names. ACE is a transport encoding.
To support the future standard of non-ASCII characters in e-mail, the
labels in DNS that are used for e-mail must handle non-ASCII characters.
You could have the e-mail group define a "E-nail ASCII Compatible
Encoding"
but that is really stupid to have in DNS. Thare should only be ONE way
to encode non-ASCII on top of ASCII in DNS. There are already to many
different ways to encode non-ASCII on top of ASCII and it is a very
difficult thing to handle in software.
Therefore we should have ONE standard for all labels in DNS.


>>  - Encoding into ACE must use Punycode WITH case marking so that
>>    case can be restored when using ToUnicode.
>> 
>> ToUnicode is fine, but decoding Punycode must restore case.
>
>I have already argued for mixed-case support as best I can.  I couldn't
>quite get it *mentioned* in the IDNA and Nameprep documents.  You don't
>have a prayer of getting it *required*.
>
>Of course, the working group that internationalizes email local parts
>would have the option of writing a spec that requires Punycode with
>mixed-case support.

Yes, that is one of the reasons why case should be preserved during
ACE conversion in DNS. To suppport the systems that require
case preservation.

   Dan