RE: [idn] Versions of Nameprep

[Hideyo Imazu <hideyo.imazu@msdw.com>]

At 00/12/13 20:43 -0800, Patrik Faltstrom wrote:
>I want people in the IDN wg think about what I wrote 
>regarding behaviour of clients and servers which use different 
>versions of NamePrep (because this will happen).

Do servers do nameprep?  According to draft-ietf-idn-nameprep-00.txt, 
nameprep is only for clients.

---begin excerpt---
Part of the chart in section 1.4 of [IDNReq] looks like this:

+---------------+
| Application   |
+---------------+
      |  Application service interface
      |  For ex. GethostbyXXXX interface
+---------------+
| Resolver      |
+---------------+
      |     <-----   DNS service interface
+-------------------------------------------+
 
In this specification, the name preparation is done in the resolver,
---end excerpt---

This is the case if DNS service interface employs UTF-8 or ACE.
Besides, this is also the case with draft-ietf-idn-idna-00.txt.

---begin excerpt---
2.1 Interfaces between DNS components in IDNA

The interfaces in IDNA can be represented pictorially as:

                 +------+
                 | User |
                 +------+
                     ^
                     |   Input and display: local interface methods
                     |   (pen, keyboard, glowing phosphorus, ...)
    +--------------- v -------------------------------+
    |         +-------------+                         |
    |         | Application |                         | End system
    |         +-------------+
    |                ^
    |                |   API call and return: nameprepped RACE
    |                v
    |          +----------+
    |          | Resolver |
    |          +----------+                           |
    +----------------^--------------------------------+
                     |   DNS query and response: nameprepped RACE
                     v
              +-------------+
              | DNS servers |
              +-------------+
---end excerpt---

Regards,
Hideyo Imazu.