Valid/invalid Label

[Alireza Saleh]

Hi,

For the past few days, I tried to answer a fundamental question to 
myself which was, "in what basis a label should be considered invalid or 
valid under IDNA documents ?".
What I currently think as the reason for valid/invalid categorizing is 
having confusion free international _domain names_ . However, I think it 
is not possible to have extra-required restrictions only on labels and 
then expect to have confusion free URI/IRI. 

Another thing which passed in my mind was that, why IDNA has been 
introduced ?  (1) This is just because of having  fancy domain names or 
(2) enabling  communities  to benefit from their languages in Internet. 
If (1), then mixing different script labels is possible and then the 
global regulation to have confusion free IDNA is required. but  if we 
consider (2) (which I think was the main reason of IDNA ), then it 
cannot be regulated globally because languages are based on scripts and 
the scripts are demonstrated by a part of the UNICODE table which I 
think it doesn't  have global considerations and its divided to 
different sections to demonstrate different scripts.

I think that a possible solution would be considering virtual links 
between one or some sections of Unicode and one or some TLDs.  Then the 
TLD operators based on those sections  which are linked to his TLD can 
make proper decisions. In other word,  TLDs would be responsible for 
their domain names. 

Another inappropriate solution might be having two different policies 
one for gTLDs and the other for ccTLDs, and of course different prefixes 
for each policy.


I'm sure taking a proper policy by all  TLDs that currently registered 
backward incompatible names between IDNA2003 and IDNA2008 both in 
character definitions or policy changes  ( such as final 
Sigma,Esszett,ZWNJ) can solve the transition problem and at least be 
sure the future incompatibilities are their responsibilities and will 
not cause any global impacts.

Best
Alireza