[soac-newgtldapsup-wg] On the IDN criteria

[Eric Brunner-Williams <ebw@xxxxxxxxxxxxxxxxxxxx>]

Colleagues,

I'm confused by the choice of metrics. We have "presence on the web", 
which could be measured by numbers of pages in a script measured by a 
convenient means, and "[number of] users", measured presumably by 
other means.
Additionally, the restriction to scripts, rather than languages, like 
the ccTLD IDN FastTrack, eliminates almost all of the Americas, and 
most of Africa, and when Han (Chinese), Arabic (Arabic, Farsi/Dari, 
Urdu), Cyrillic and Indic scripts are eliminated (see either metric 
above), leaves very little likely to be of use.
I don't think the proper scope of using the DNS as a means for 
languages communities to offer persistent resource identifiers to the 
readers, writers and speakers of those languages should be limited to 
exclude those languages which have managed to survive the imposition 
of Latin, Cyrillic, Han, or Arabic orthographies.
Restated, our mission is not to encourage the destruction of living 
languages currently encoded for convenience in Latin, Cyrillic, Han, 
or Arabic scripts.
I suggest the better choice is to use "language" rather than "script", 
as the element of diversity, and of public interest, to be advanced.
There are several languages that use more than one script. Yiddish is 
written in either Hebrew or Latin script. Arabic is written in either 
Arabic or Latin script. Hindi-Urdu is written as Hindi in Devanagari 
script, and as Urdu in Arabic script. Cree is written in either 
Unified Canadian Aboriginal Syllabary or Latin script. Cherokee is 
written in either Cherokee Syllabary or Latin script ...
There are scripts which are used to write more than one language. Many 
Indo-European languages are written in Latin script. Urdu, Dari, 
Farsi, and Arabic are written in Arabic script. ...
The 20% rule usefully addresses a language community using two 
scripts, and a community using two or more languages.
The parenthetical note on the next bullet item, "An NGO, company or 
other entity  ..." I don't understand. What technical support could be 
offered and why would the applicant's technical needs be 
distinguishable from any other applicant's technical needs? 
Clarification of what is meant by "technical support" might remove 
this lack of clarity.
For competition policy reasons I would prefer that the forth bullet 
item contain an explicit limitation so that Verisign can not claim 
"support" under any color of serving the public interest.
The use only one script rule is particularly unfortunate. It is 
irrational to apply for only a Unified Canadian Aboriginal Syllabary 
string, as more than half the populations speaking languages encoded 
in UCAS in the Northern Roadless Area, live south of the Northern 
Roadless Area and use Latin script. It is even less rational to apply 
for a Cherokee Syllabary string, as the majority of Cherokees (all 
three polities) use Latin script. Yet in both cases, the common public 
policy goals of Aboriginal Language Maintenance and Recovery are for 
the dual use of both scripts, not the abandonment of one, and the loss 
of its speakers to the dominant language, English or French.
A similar situation arises in Sub-Saharan Africa, where Latin and 
Arabic scripts co-exist.
If we were sure that an application for ".foo" in Script X would, 
within 12 or 24 months, receive a delegation for the corresponding 
".bar" in Script Y, then there would be a way to work around the 
unfortunate effect of selecting the largest (number of domain buyers) 
language/script market first. Absent this, the pick the most money 
approach means we will have created discrimination within communities, 
and therefore be responsible for the strife that may follow.
I look forward to tomorrow's concall and discussion.

Eric