Re: Standardizing on IDNA 2003 in the URL Standard

At 02:39 30/01/2014, Anne van Kesteren wrote:
>However, I think I have been convinced by this thread that UTS #46
>might be good enough as replacement for IDNA2003. Once it has been
>clarified per the feedback I submitted I will incorporate it in the
>URL Standard. It's unfortunate that even #46 is implemented in
>different ways. :-(


I am interested in the pragmatic "InterUse" aspects that I need (cf. 
infra). If you bear with me a minute, I will explain how I understand 
where the whole system stands, how it is probably evolving, and where 
the practical solution should be or where you could locate them and 
how, with who it will have to be discussed, and self organization 
will propagate.


The internet has an architectonical doctrine (IEN 48), architectural 
plan (IAB), technological author (IETF), documentary embodiment 
(RFCs), practical operation guidance (BCPs), registries (RIRs), 
operators (ISPs), enabling protocols (code) in order to answer its 
motivations (or more if possible).

The author of IEN 48 is the chair of this WG. We, therefore, have 
through him a remarkable continuity. I see in these texts (IEN 48 and 
the WG of this charter)  - 
- a remarkably simple, robust, and efficient logic that I qualify as:

- "internet": proof of Louis Pouzin's catenet concept on a global 
plane (i.e. non-formatted ASCII text oriented interconnectivity and 
access). Completed.

- "intertech": exploration of a multitechnology support by this proof 
of concept, i.e. the addition of the layer six presentation/Tymnet 
equivalent. Missing, but many uncoupled applications (e.g. Web).

- "interuse": "an improved means to produce and use stable and 
unambiguous IDN identifiers" for everyone (network access + machine 
technology + people culture). (cf. RFC 3271).


The problem of this WG was to address the interuse stratum without 
the intertech stratum having been coherently documented, tested, and 
proven. This is why:

2.1. initially I said that I would support the WG work on the 
internet stratum in order to support my own additions that I called 
the ML-DNS (ML having multiple meanings corresponding to various 
multilateral, multiledger, multilayer, multilingual, etc. aspects).

2.2. I opposed Vint Cerf when he attempted a layer violation in 
trying to incorporate layer six tasks in the end to end internet 
architecture side.

2.3. This was elegantly addressed by Pete Resnick and Paul Hoffman's 
future RFC 5895 which "unusually" exemplified in an IETF layers 
document how the intertech strata could interface the internet 
specifications and requirement. Actually it meant that no fringe to 
fringe (cf. RFC 1958) MUST was incorporated in an end to end RFC.

Today, the problems we face are the practical interuse problems while 
the intertech is still in limbo as a set of post-presentation systems 
(such as the Web).


I was reluctant to work on this intertech stratum because these 
things demand time (and money) before being commonly understood, 
worked on, and tested, etc. The huge risk was to be passed by crime 
and or merchants (RFC 3869) because a quick and *dirty* solution 
calls for very limited work. The difficulty was to avoid the dirty 
aspect, hence to create a forum for everyone to be able to work on a 
matter which is similar, in continuation, but foreign to the IETF, 
and to make sure that there was no area conflict with the IETF and an 
established bridge (now IUCG@IETF). The intertech means providing a 
presentation layer six (PL6) as a "plugged layer on the user side" 
(PLUS) as part of IUIs (intelligent use interfaces).


The Post-Snowden general awareness removes the risk of the crime but 
increases the merchants' risk (moreover, Vint is now on Google's side 
with Harald Alvestrand and Mark Davis). The well-organized OpenStand 
+ /1NET operation shows it. This distracts people from the real 
concern: all the NSAs of the world operate easily on a very 
vulnerable Internet, and so the final distrust is in the IETF 
capacity to produce a robustly secure technology.


Now, please remember that I come from Tymnet Extended Services, i.e. 
what followed in the "internet pile" over theinternet, intertech, and 
partly interuse, i.e. is what I call the Intersem, the semiotics 
stratum, the internet of ideas/subjects. There, I have two problems 
that are being quite discussed right now:

- the ONS, i.e. the integration of the IoT, internet of things, into 
the picture. I am interested in understanding how Fadi Chehehade, 
Mathieu Weil, GS1, etc. are going to proceed and how this will 
compare with the IDNS in its relation to the DNS Question: will this 
help/hamper the general semantic addressing system and its digital 
names as we plan it (strict conformity to the DNS, as IDNA with the 
addition of CNAME confusion for class support).

- the pseudo "balkanization" of the internet that is a commercial 
campaign for the UNICODE "globalization/internationalization" against 
the "nationalizations" of WCIT US opponents which is a risk to 
internet stability resulting from the support of the IANA by a 
structure with political ambitions.

Time has, therefore, come to satisfy everyone through the 
globalization of the IANA (ICANN has no real technical interest) 
within a multistakeholderist framework. The fragmentation is linked 
to the dispute over the root file. The solution is, therefore, to 
come back to the initial Tymnet days and suppress the fragmentation 
risks in making the top zone data ubiquitous. This is the very simple 
HomeRoot project.


This means empowering everyone with his/her own root file, so 
everyone is a multistakeholder on an equal footing. This is the 
"HomeRoot" project.

1. as long as the data are the same there is no misdirection risk.
2. since the resolution is carried at the user fringe, there is no 
buffer cross pollution. Anyway, TTL decrease with AJAX.
3. the root servers will remain in an ever increasing traffic, so 
that the NSA will keep some metadata processing .
4. this is a clean technical response to ICE that does not not 
require national root server systems.
5. IDNA2008 modules and value added modules (supporting IDNA2003, 
etc. variations) will be possible (for example, through IDNAxxxx 
transition classes).
6. it will be possible to fully use the class mechanism for a unique 
internet with multiple visions.
7. the resolver will serve as a seed for the IUI (Intelligent Use 
Interface) to support new services, technologies, network 
applications, OPESes, etc.

An MS-IG debate (or at least parts of it that IUsers would support) 
could advocate:

- an integrated installation and parametering the support of a local 
resolver on Windows/Linux/Mac
- the publication of hint files by trusted third parties and governments
- the architectural definition of the IUI within an interpresentation 
framework (netix) extending the posix system interface to network use.
- in order to allow the definition, documentation, and deployment of 
format presentation services local operating tasks within the IUI 
Interbox transparently to the user application.

You will note that this is in pure thinking continuity with IDNA where:

- the user enters the "data" of an identifier in their own way.
- that are treated by IDNA to become the identifier "tracta" sent to the DNS
- that become the "capta" received by the other side
- that transformed them into final "tracta" that the application will 
be able to use (which may differ from the initial "data" if the IDNA 
context is not the same).


This corresponds to an additional technological and thinking plane, 
i.e. in addition to information and communications, to the 
intellition plane and to ibits, i.e. intelligently implied bits. An 
example is what PRISM delivers as a pertinent reality information 
system monitor. The transiting data are possibly treated by OPES on 
the wire. And that OPES can be interconnected in underlying ONES 
(open network of edge systems). Just to stay in the IDNA context, 
ONES may coordinate the treatment of variants in adding syllodata to 
the datagram/domain name metadata (i.e. the data between the data, 
for example between linked data).

I fully understand that these considerations are new to most. They 
just belong to the continuation of the agoric approach initiated by 
Norman Hardy the conceiver of the Tymnet architecture. Agorics is the 
polylectic (networked) generalization of Aristotle's logic dialectic 
(x and y then z conclusion) and of the Wiener cybernetic monolectic 
(action then reaction). Agorics results in emergences related to the 
networked syllodatas attractor.

I suppose that the MS-IG will soon look seriously at these notions, 
spurred by projects such as HomeRoot, PL6 and DARPA cyber grand 
challenge. This is in line with the INTLNET MDRS project (metadata 
referential/multilingual distributed registry system/service) that I 
introduced at the 2006 UNESCO/ITU meeting as a continuation of the 
INTLFILE that became the root file. An agoric consideration of the 
DARPA CGC shows that real-time tracta correction (i.e. a super 
semantic firewall) will demand it to encapsulate an architectony, 
i.e. a semantic general model of the universe that is able to 
understand the context and the pertinence of the inputs. This kind of 
system should be able to correct the IDN related uncertainties.


I suspect that the political and societal governance evolution will 
probably pair with the technical evolution, so it will probably be an 
interactive process once users understand that they can interact with 
the technical development and that it depends on no one they could 
distrust. Distrust will then be only of themselves: they will 
consider their involvement as dangerous until some experience and 
press show and explain to them that this is not complex. A common 
e-literacy issue.

This is why technical governance is of the essence: it will have to 
influence the coordination of the different initiatives and uses of 
the internet as a test-bed (cf. ICANN/ICP-3).

1. improvement and simplification of BIND installation on Windows 
including default parameter files.
2. development and support of an IANA root+ local update tool.

3. definition of a network oriented open code license.
4. split of a browser as a virtual machine acting as an IUI + screen.
5. documentation of the parametering and of the interaction between 
IUI (presentation layer 6 project).

This means, in this case, that users and user applications could 
forget about scripts that will be handled by the IUI OPES that will 
punycode encode/decode the data/capta flow as part of the layer 6 
functions. Synchronization issues if any should then be dealt with a 
MS enhanced cooperation to be formed among IUse actors, IANA, IETF, 
UNICODE, W3C, etc.

Received on Thursday, 30 January 2014 18:35:02 UTC