[idn] Re: Alternative Solutions

["Edmon" <edmon@neteka.com>]

> At 05:23 PM 5/1/2001, Edmon wrote:
> >Server-side ACE - ACE transformation takes place at server end
>
> If the transformation is done at the server, then the data travel from the
> client as binary.  How is that different from your second alternative?
>
> For that matter, why bother with an ACE transformation at that point?
>
> Please clarify.

There are a number of reasons, including:
- ACE domain can be used after the DNS, such as web & mail (part of the
argument for using ACE)
- ACE domain could be used for administration purposes, for sys admins to
configure and edit zone files, etc.
- will be able to transition to the "long-term" protocol extension solution
easily

>
>
> >Protocol Extension - use extension mechanisms such as EDNS or other bit
> >flagging & use UTF
>
> This is a class, "dual-stack" operation.  One stack is the original system
> and the second is the new.
>
> (By way of raising a warning flag early, one danger with this approach is
> what we have seen with IPv6.  Namely, people will try to add zillions of
> changes and features, thereby delaying the effort by years.)

EDNS is pretty much in place and other extension mechanisms could be easily
introduced also.  As to lengthy discussions, it depends on whether we are
zoomed into providing a multilingual dns only and leave bells and whistles
to the implementors.  So long as the core platform is there, if implementors
decide to add features and delay their own roll out then its their decision.

Edmon