referencing IDNA2008 (and IDNA2003?)
J-F C. Morfin
jfc at morfin.org
Fri Oct 22 13:12:59 CEST 2010
what IDNA2008 did was to determine the Internet border with the
users' area and to build an Internet Iron Curtain on the Internet
side guaranteeing the independent DNS stability for ever. I was one
of the most concerned by the location of the future curtains on
behalf the users' interests and coherence with other technologies.
Because up to now new diversity was supported by technology internal
growth and will now have to be supported by external multiplicty.
The concern was over the UTS 36 and RFC 5895 like issues and
presentation layer, orthotypography, semantic addressing, user side
architecture coherence, fringe intelligence, multilinguisation, etc.
related issues: should they be dealt with inside or outside of the
Curtain. I raised the question from the very onset, asking the Chair
how he understood the Charter. The response was by both Chairs - of
the initial WG/IDNA (IDNA2003) and of the WG/IDNABIS (IDNA2008) -
these issues were mostly outside of the IDNA charter, provided
IDNA2003 could be accommodated as much as possible.
I then committed to address these issues through a multi-layer
encapsulation of the DNS on the user side (ML-DNS). This is what I am
now working on providing an extended Linux/Windows end user solution
to be first experimented by the Internet Users community. After I
made (through appeals to the IESG and IAB) made sure that Internet
network legacy and Internet Use where considered as two different
standardization areas that currently liaise through the iucg at ietf.org
The remaining problem was to clarify what "mostly outside of the IDNA
charter" meant, i.e. to define precisely where was the border. The
border was eventually consensually agreed at lowercase A-labels. I
reported the IESG that as promoter of the Project.FRA (Francophone
space) and of the Intersem (semiotic Internet of thoughts) :
1. I was satisfied with this location, and urged them to approve the
2. but that it led me to address several key-issues at a "no-mans
land" between the Iron Curtain and the users.
2.1. One of the typical usage problem was orthotypography, when
Unicode does not specifically support some metadata necessary to a
proper semantic presentation like in the case of French majuscules,
they support in using ASCII uppercases. This will be addressed by the
ML-DNS and a transparent extension of the punycode algorithm (I
published a draft to discuss possible solutions)
2.2. One of the typical architectural problem is the IDNA
(Internationalized domain names in applications) architecture itself
which permits different user applications to behave differently and
to resolve different addresses for the same U-label. This point was
also raised by the AD : it was not specific of IDNA2008 but is one of
the issues the IAB has on its shopping list (cf. John Klensin Draft
on IDNA). My solution is that the "no-mans land" becomes part of the
Internet Use architecture an "Internet Use Interface [IUI]" and among
many other extended services to diversity and innovation provides an
unique IDNA (Internet Domain Name Application) service to the users
and to the Internet DNS (and other DNS) for the users to uniquely resolve.
IMHO in the cookie case, there are three types of cookies:
- those documented as part of the Internet technology, subject to
standard track document, that can only document their use within the
Internet Iron Curtain (i.e. lowercase A-labels)
- those being used by IUI level applications that are to be
documented by the Internet Users community that can be documented
through BCP or for information like RFC 5895.
- those being used by the users applications that are documented by
The IU community has many needs and solutions that can be supported
in respecting the now existing and confirmed Iron Curtain separation.
Let try to not erode it and create instability.
To make sure everyone understand IDNA2008 is here to stay, we call
IDNA2010 the debate on the IDNA2008 support on the User Side, and
IDNA2012 the debate of the mutual support of IDNA2008 and IDNA2010.
More information about the Idna-update