Proposed document structure and content for IDNA

Michel SUIGNARD Michel at suignard.com
Fri Jul 11 15:59:21 CEST 2008


>From Frank Ellermann 
>>JFC Morfin wrote:
>>
>> one thing: to use "ISO 10646" than "Unicode".
>
>The "tables" and "BiDi" need Unicode properties
>defined in TUS.  AFAIK (I never looked into it)
>the other memo doesn't specify these properties.

It should be noted that ISO 10646 by its normative referencing to the
Unicode bidi algorithm and normalization is in fact equivalent to
Unicode for most of the IDN work. However some other properties such as
General Category, identifiers, etc... are not yet. There is work to
better align Unicode and 10646 in the next edition of 10646 by
increasing normative references between the two standards. At this point
there is no need to try to create a wedge between 10646 and Unicode. If
your affiliation makes it easier to work on the ISO context, for example
through ISO/IEC JTC1/SC2 and SC2/WG2, you get the same benefit and a
very similar influence to the result of the work between 10646 and
Unicode. Liaisons to Unicode are also very effective at influencing the
resulting work.

In other words, we should not change the referencing to Unicode in IDNA.
There is another reason which is the typical delay created by using ISO
10646 (typically ISO amendments are officially published 6 months or
more later than the equivalent Unicode version).

Finally, I would also like a simpler structure for IDNA 2008, but can
live with the current one, especially if the merge significantly delays
the outcome. 

Michel



More information about the Idna-update mailing list