[vip] Draft on IDN Tables in XML

[JFC Morfin]

At 07:30 12/03/2012, Abdulrahman I. ALGhadir wrote:
>Hello,
>Is it possible to add the relation type for each variant like is it 
>exactly similar or partially?
>Also, what about the registry rules? Some registries impose more 
>restriction on the IDNA protocol is it possible to add a tag the describe them?
>AbdulRahman,

There is a need for a "netlocale" file, describing the network 
multilinguistics (i.e. what is related to the mutual support of the 
linguistic diversity) as seen from a user (i.e. in the languages of 
the users) and operated by his/her linguistic tools.

Until now the BCP47 has acknowledged three *main* parameters 
(langtags) which are language (ISO 639), script (ISO 15924), and 
"legal culture" (ISO 3166). One  of the consequences of IDNA2008 has 
been to consolidate the powerful but not yet documented Internet 
Presentation Layer and to introduce the need to extend the BCP47. 
This also create concerns for the WG/PRECIS and the Happiana list. 
Vint Cerf, as the Chair of the WG/IDNABis suggested that the 
post-IDNA2008 work would be coordinated by ICANN. As IUsers 
(Intelligent Internet Independent users) we could not accept this 
suggestion until ICANN and IETF would have defined their position in 
the resulting technology stewardship.

VIP is a preliminary step, as is Google+, as are the Drafts I 
published. All this leads us to the addition of at least three new 
parameters to BCP47, a new rule set and a coordination authority:

- parameters are: registry (i.e. root name - the TLD for what 
concerns the Internet), CLASS, and presentation. TLD are used and 
known, but have to be extended to non-Internet needs (IRT the whole 
digital ecosystem administration). There are 65.536 CLASSes, only one 
being actually used, and ICANN (ICP-3) pushing new needs to the use 
of more. Presentations are not yet defined, but we know one as 
related to "xn--" and punycode.

- new rule set: this six main parameters have to be documented 
through a polylingual (i.e. a 7th main parameter) format, that will 
have to be able to support different set of rules (by TLD manager, 
about variants, by local Laws, etc.)

- a coordination authority. All this affects the daily operation of 
browsers, applications, law enforcement (for example ice.com in the 
USA), new intellectual protection rules (e.g. the "babel names" which 
have not even been alluded to in the VIP case: when all the 
characters of a domain names are not supported, and in some 
applications/browsers for security reasons, the U-label may be 
replaced or completed by the A-label and the A-label to display a IP 
protected name. With additional problem with variants, since the 
several A-labels may correspond to several U-labels).

For obvious security and operational reasons, all this information 
should be at least moderately (within a TTL like delay) synced to 
avoid blocking situations, which may add to not considered non tested 
possible conflicts with IPv6 and DNSSEC.

We have now to consider the figures:

- languages (ISO 639-7), probably 25.000 language units. Beware: in 
our world, languages will be considered as high level common 
communications protocols between "semantic processors", natural 
(human brain) and mechanical, hence as "mecalanguages". There will be 
advanced variants, revived old languages, encrypted languages.

- scripts (ISO 15924), 100+ scripts, with in addition signs and 
possibly accents (natural or binary).

- countries and equivalent (ISO 3166 includes international bodies' code).

- TLD, impossible to determine as the New gTLD Program in the 
ICANN/NTIA "IN" CLASS is theoretically unlimited.

- CLASSes : 65.536 (reminder: a CLASS is a way of perceiving the 
Internet protocols and DNS)

- Presentations : the number is be essence unlimited, however one may 
expect two kinds of presentations : externets and private. An 
externet is an "external network look-alike" built in using the 
Internet. One can also describe them as "open closed garden": an 
Internet subset where procedures, operational constraints and 
services, rates, neutrality, languages, etc. are differently 
supported. The common documentation should probably only concern the 
public externet system. Private presentations includes special 
languages, interface uses, encryptions, They should obey to the same 
format, but will most probably not be synced with all the other BCP47+ data.

- DNS visions may also be added.

- and then multiply all this by the number of languages again to be 
presented in every language and scripts (some languages being 
supported by several scripts).

jfc