[arabic-vip] Initial list of issues and questions

[Siavash Shahshahani]

Regarding the presence of digits in TLD I got conflicting results:
1. I asked a couple of registry operators, both said that the only
restriction (for ASCII) is that the TLD not start with a digit. But this
may be from the older version guidebook because:
2. In the new Guidebook(redlined version), Module 2, p.2-11, article
1.2.1, it says that ASCII label must consist entirely of letters. Also on
page 2-12, article 2.1.5, it says that all characters within the label must
have the same directional property (thus eliminating Hindi-Arabic digits
which are L-to-R within the R-to-L script).
Maybe Baher could ask Kurt Pritz for an authoritative clarification.
Siavash 



On Sun, 19 Jun 2011 09:50:46 -0700, "Dr.Sarmad Hussain"
<sarmad at cantab.net>
wrote:
> Dear All,
> 
> As per the discussion in the meeting on 18th June, we met again on 19th
> June
> to discuss the issues.  Here is a summary of our initial discussion for
> further feedback:
> 
> 1.  Extra normalization - composed and decomposed forms not considered
> equivalent by Unicode
> 2. Optional combining marks - aerab (e.g. fatha, damma, kasra, shadda,
> etc.)
> 3. Conflated combining marks - alif, vao, yeh, hah with and without
hamza;
> alif with and without madda, etc.
> 4. Two characters/unicodes with same shape in a particular position
> (initial, medial, final or isolated), e.g. arabic and persian kaf; yay;
> etc.
> 5. Two characters/unicodes with similar shape; e.g. swash and other kaf;
> tay
> marbuta and hay, etc.
> 6. ZWNJ with characters which do not change shape perceptibly
> 7. mechanisms for documenting these cases for eventual comparisons to
> determine variant TLDs
> 
> Following issues were pended for consideration until the associated
> questions are responded by ICANN staff and consultants:
> 
> 1. Confusion caused by digits - Question: Though digits are relevant for
> labels, are they relevant issue for TLDs?
> 2. Specification of "language" table - Question: The way the table is
> defined in the documents circulated by ICANN in the context of IDN
Variants
> project may not be relevant for Arabic, but should the team consider
> defining the proper table format an issue? How is the concept of table
> relevant for TLDs (as it is normally not used for catering to labels at
the
> top level).
> 3. Blocking, reserving, etc. are strategies being discussed in the
> documents
> circulated by ICANN. Question: How are they related to the variant
issues?
>  Are these strategies part of the problem or part of the solution for
> variants?  If latter, should they be discussed at this stage of the
project
> where only issues are being considered?
> 
> The following sources of variations would not be considered:
> 
> 1. Homophones
> 2. Bidirectional issues (as they are not relevant for a TLD, but to the
> complete domain name)
> 3. writing style - as this is largely a font issue
> 
> 
> We look forward to the feedback on all the items.  We request feedback
on
> the second set of questions by ICANN staff and consultants.
> 
> regards,
> Sarmad