New version, draft-faltstrom-idnabis-tables-02.txt, available

[John C Klensin]

--On Tuesday, June 12, 2007 15:53 +0200 JFC Morfin 
<jefsey at jefsey.com> wrote:

> Dear Harald,
> you obviously measure the political impact of such a
> restriction not being very precisely documented.
> IDNA made a distinction between countries on the ASCII TLD
> basis. IMHO, if IDNAbis only enlarges the basis of this split,
> the IGF will simply conclude that the global approach is
> inadequate and call for another architecture.

Jefsey,

Regardless of anything else that is going on here, IDNA does not 
make any distinctions at all based on TLDs, ASCII or otherwise. 
It is a fairly fundamental consequence of the DNS design and 
architecture (having nothing specific to do with IDNs) that 
storage and lookup (matching) algorithms for fully-qualified 
names and the labels that make them up cannot depend on location 
in the hierarchy -- either depth or subtrees.  Any 
implementation of the DNS that assumes knowledge of the 
language, script, or other properties of a given label based on 
the label (TLD) at the top of its particular subtree is 
non-conforming and almost certainly will not interoperate well 
with other DNS implementations.

Independent of whether they would be a good idea or can be 
enforced, one can make registration policy decisions about what 
strings can be registered in particular subtrees and under what 
conditions.   But those restrictions have nothing to do with 
IDNA (or IDNAbis) themselves -- the protocols just deal with 
strings and labels.

        john