Re: [idn] ToUnicode output can be longer than input

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

Adam M. Costello wrote:

>> There are several domain names that will fail when ToASCII is used,
>> but are still domain names.  They just cannot be handled by IDNA.
>
>There are non-text domain labels to which ToASCII cannot even be
>applied, because ToASCII can be applied only to labels that are text.
>But among text labels, ToASCII defines which ones are valid.

Valid for IDNA, not valid in general.

>
>> IDNA do not define the world and do not define the basic semantics
>> of ACE or domain names with non-ASCII characters.  IDNA does only
>> define a way to encode domain names so they can be sent over lagacy
>> ASCII DNS protocol.  It does not define what domain names work in an
>> international context.
>
>The IDNA spec disagrees.  It says:
>
>    This document defines internationalized domain names (IDNs)...
>
>    If an application wants to use non-ASCII characters in domain names,
>    IDNA is the only currently-defined option.
>
>    Applications can also define protocols and interfaces that support
>    IDNs directly using non-ASCII representations.  IDNA does not
>    prescribe any particular representation for new protocols, but it
>    still defines which names are valid and how they are compared.

Maybe IDNs, as that is a construct of IDNA, but not domain names
in general. Domain names in an international context is not defined
by IDNA, IDNA have limitations so that it cannot handle all
domain names that can exist in an international context.
I can see no reason to limit the international world due to limits
of ASCII and the solution selected by IDNA.

When an international context of the DNS protocol is defined, it can
support more domain names than can be handled by IDNA.

   Dan