[arabic-vip] Issues Document

[Siavash Shahshahani]

Hello Francisco,
My personal impression is that the Arabic-script group is also on that
same wavelegth. The exchange below was later continued, and Sarmad's latest
version of Issues Document should reflect that. Sarmad admits that
'language table' is a misnomer and I have been pushing for avoiding the
term 'language table' because of the implications it may suggest. More
neutral terms such as 'Registry Table' or 'TLD Table' can be proposed. I
personally think 'TLD Table' is more accurate because the same registry may
be running several TLDs and using different subsets of say the full Arabic
script for different TLDs. On the other hand, each TLD, cc or g, must have
a well-defined table associated with it. My impression is that Sarmad was
just pointing out that managing tables associated with a limited and
well-defined set of languages may be easier. This may be true, because so
far the IDN experience, limited to fast track, has concentrated on
languages and single nations. However, to have thriving gTLDs in IDN, one
must go beyond the language and national barriers. So it seems to me that
the general trend you are noticing in various groups is the correct trend,
as long as one understands that not all potential variant cases identified
per complete script table are necessarily applicable to all TLD tables
constructed from a subset.
Regards,
Siavash

On Tue, 6 Sep 2011 13:49:40 -0700, Francisco Arias
<francisco.arias at icann.org> wrote:
> Hello,
> 
> There is one issue raised in the draft outline/list of issues circulated
> by ICANN related to this. The issue is whether the Variant Tables in the
> root should be language or script based.
> 
> From discussions in various teams it seems like the best solution would
be
> to have a table per script in the root, not by language. The reason
being
> you have the potential to have conflicting Tables if they are based on
> language, and given that in the root there is nothing to infer the
> language (vs. a ccTLD for example).
> 
> Is the Arabic team proposing to have Variant Tables based on language
for
> the root?, I'm not sure how to interpret the text below.
> 
> __
> Francisco
> 
> 
> 
> 
> 
> On 9/4/11 11:44 PM, "Sarmad Hussain" <sarmad.hussain at kics.edu.pk> wrote:
> 
>>Agree that "language" table is a misnomer.  Not sure what to call it,
>>perhaps a "registry supported character set and variant table"?
>>
>>I am rewording this in the revised version 0.4 (will circulate it
>>tonight).  Please check the text and make sure it meets your
expectations.
>>
>>Regards,
>>Sarmad
>>
>>-----Original Message-----
>>From: Siavash Shahshahani [mailto:shahshah at irnic.ir]
>>Sent: Thursday, September 01, 2011 8:58 AM
>>To: Sarmad Hussain
>>Cc: arabic-vip at icann.org
>>Subject: Re: [arabic-vip] Issues Document
>>
>>Dear Sarmad,
>>Thank you for the great combining work, well done. I wish to point out
>>something about the following:
>>----------------------------------------------------------
>>a.	Management of Language and Variant Tables
>>
>>The registry should decide what are the supported languages along with
>>defining language table and variants table for each supported language.
>>Here are some questions (about this issue) that each registry should
>>consider:
>>
>>i.	What are the supported languages in the registry's TLD?
>>..... etc
>>-------------------------------------------------------------------------
>>I have trouble with the 'should' in the introduction. Why should it? If
>>you are registering an ASCII domain you are not required to discuss the
>>language. Language is taken care of by the requirements of registration;
>>no
>>further limitations are needed. This will be very important by gTLDs.
Note
>>that a label need not be carry a meaning in any language. The only
>>sensible
>>requirement for the gTLD would be that its rules and regulations take
care
>>of variants in a way that no threat to security and stability would
ensue.
>>There is no universal solution for this; the solution would depend on
the
>>character table used by the regisry. In the case of Arabic script,
>>depending on which subset of the UNICODE table you are allowing for
>>registration, your requirements will differ; no reference to 'language'
is
>>needed.
>>Regards,
>>Siavash
>> 
>>
>>
>>On Tue, 30 Aug 2011 16:32:08 -0700, "Sarmad Hussain"
>><sarmad.hussain at kics.edu.pk> wrote:
>>> Dear All,
>>> 
>>>  
>>> 
>>> Thank you for the feedback so far.
>>> 
>>>  
>>> 
>>> I have incorporated comments from Fahd, Dr. Al-Zoman and also included
>>> change due to our final discussions on dispute resolution, and end
user.
>>
>>> 
>>>  
>>> 
>>> Please find attached an updated version for your review and further
>>> feedback.
>>> 
>>>  
>>> 
>>>  
>>> 
>>> I wish all of you a happy Eid and we will have our conference call on
>>> Tuesday, 6th Sept.
>>> 
>>>  
>>> 
>>> Regards,
>>> Sarmad
>>
>>