My comments on the report

["李光皓" <Liguanghao@xxxxxxxx>]

I appreciate the efforts that GNSO has put into this report. And I have 
comments for the following points.

First, in the recommendation two, there is a clause for “confusingly 
similar”. The report has not formally defined what it is the scope of the 
term. If it means “visually, phonetic, and conceptually”, then I would like a 
clearer definition of what is considered as conceptual confusingly similar. And 
I would like to see an example or an actual case on this.

Second, in the recommendation nineteen, it requires that the new TLD registry 
only use ICANN accredited registrar. Although we all agree that maintaining the 
stability and the safety of the DNS is in the highest priority, we also agree 
that we should foster a system that can best meet users’ requirements. Thus, 
the model of use only ICANN accredited registrars may not suitable for new TLDs 
that involves IDN strings. I see the reasons as following:
IDN TLD is targeted to serve a specific community or local area, thus it has a 
local characteristic. And the local registrars have advantages in language, 
culture, close local services and best understanding of local needs. Such a 
restrain, has setup obstacles for these registrars to serve the customers, in 
term, it hurts the popularization of the IDNs. Thus, I recommend ICANN should 
setup a series of criteria and allowing the registry to accredit more local 
registrars in compliances with these criteria. In short, I propose that the 
management system should be no less creative than the technology that brings us 
forward. 

Third, I further recommend that there should not be any “IDN version of ASCII 
gTLD”. Any IDN TLD string should be a new TLD in no relationship with ASCII 
gTLDs, on top of the IDN TLD should not be confusingly similar to the current 
TLDs. I see the reasons as following: 
1.                 there is no “correct” translation. An ASCII TLD string 
usually is an abbreviation of a word in English, and thus, it may have many 
interpretations of the same string. For example, “COM” may be interpreted as 
“commercial”, “company”, “communication” and so on. Moreover, even a 
string has a common understanding in English, when translated, it usually have 
more than one corresponding interpretations. Due to such a reason, an ASCII TLD 
is extremely difficult to have an “IDN version”.
2.                  in the aspect of user experiences, if an IDN TLD is to be 
managed by a registry which is based in English speaking countries, its related 
policy, dispute resolution and other administration will very likely to be in 
English, and utilize ADR institute base in these countries. All of these will 
create obstacles for non-English speaking users, especially in the dispute 
resolution. In turn, it will force users which the TLD string intends to serve 
to register the domain defensively instead of creating new values for them. It 
thus against the original purpose of creating IDN.

Guanghao Li