FW: [centr-tech] IDNA Redux]

[Marcos Sanz/Denic]

Vint,

> To avoid serious confusables problems that we already know have been or 
will
> be exploited, we many need to yield to unique "identifier" requirements 
even
> if this means the allowed character set does not permit precise 
"language
> oriented" expression. 

I agree. However the rationale in draft-faltstrom-idnabis-tables about 
which characters are allowed to be an identifier or not is not very clear 
to my eyes. What about taking UAX #31 as a starting point and then 
refining on that, if need be?

Best regards
Marcos