Re: [idn-discuss] Some Questions About IDN

[John C Klensin <john+icann@xxxxxxx>]

--On Thursday, 31 May, 2007 18:05 +0500 Muhammad Farooq-i-Azam
<i80x86@xxxxxxxxx> wrote:

> Hello All!
> 
> I am part of this list for quite some time now. Recent mails
> and
> traffic on this list
> has prompted me re-kindle my interest in IDN. I am particulary
> interested to
> deploy and introduce IDN in Pakistan.
> 
> I have some questions concerning IDN which I hope experts on
> this list would mind answering:
> 
> 1- I have searced the web but could not confirm that IDN has
> been implemented in
> any widely used DNS servers like BIND. Can anyone confirm is
> this correct?

IDNs do not require, or, if the Standard is followed, even
permit implementation in a DNS server.  So the answer to your
question is that no Standards-conforming server, widely used or
otherwise, has implemented IDNs, nor is that in any way
necessary.

> 2- What about browsers, like Firefox or IE? Is there any
> support available in these browsers at present?

Support is available, either directly or via plug-ins, in all
known popular browsers.  Firefor and IE7 support IDNs directly
although both have made provisions to protect users from some
IDN side-effects that you might or might not consider helpful.
Those provisions and approaches are quite different between the
two.

[Tina, I wonder if it would be worthwhile for ICANN to start
compiling a list, and perhaps even reviews, of conforming
implementations, both in applications (browsers, email clients,
etc.) and code libraries?] 

> 3- Now an important question for me.
> Apart from the above implementations of IDNA in client and
> server, is there
> any other requirement that a non-english language needs to
> fulfill to
> start using
> Internationalized Domain Names? In particular, I have read
> somewhere in the RFCs
> about "Language-Specific Character Tables" which are to be
> developed prior to
> the usage of a language for IDN. *******What are these
> character tables?******
> My language URDU is almost completely represented in UNICODE
> so that
> even websites with URDU unicode text are now available. Again,
> my question is:
> Is there something else also needs to be done apart from
> client and server side
> implementation of IDNA e.g. development of
> "Language-Specific Characater Tables".
> If so, what are these character tables.

There are at least three things about which you should, at
least, be aware.   In no particular order:

(1) At least as I understand it, current ICANN guidelines
require (for gTLDs) and strongly suggest (for ccTLDs) that
"language-specific character tables" be registered for each
domain to identify the characters it believes are associated
with a language it finds of interest.   For Urdu, the intention
would be to have a list of the subset of Arabic and
Arabic-derived characters that are used to write the language,
i.e., the definition of the Urdu script as your domain
understands it.  (See also below.)

(2) There are some problems in the handling of scripts that are
written from right to left in the current (2003) version of the
IDN Standards (IDNA and associated documents).  Those issues are
discussed in a working document,
draft-alvestrand-idna-bidi-00.txt (you may need to do a Google
search for this -- it is officially "expired" pending posting of
a new version).  You might want to take a look at that
discussion before proceeding.  If you have comments, we would
very much like to hear from you (see below).

(3) More generally, an analysis was performed of perceived
difficulties with the existing versions of the IDN Standards and
published under the auspices of the Internet Architecture Board
(IAB).  That document can be found at
ftp://ftp.rfc-editor.org/in-notes/rfc4690.txt (and elsewhere).
It has been followed up by technical work to produce a more
workable version of the standards. The document about right to
left ("bidi") scripts mentioned above is part of that effort.
The overview document is draft-klensin-idnabis-issues-01.txt
(available at
http://www.ietf.org/internet-drafts/draft-klensin-idnabis-issues-01.txt
and elsewhere).   That work is being discussed on a mailing list
specific to the IDNA revision effort (not to more general
discussion or related issues, for which this list is
appropriate) on which we would welcome your comments and
participation after you have looked at the relevant documents.
Information on that list is available in the documents
referenced above.

regards and best wishes,
   John Klensin