[discuss] So-called alternate roots

Michel Gauthier mg at telepresse.com
Sat Jan 4 16:52:43 UTC 2014


I am only an analyst who professionnally knows enough things about 
computers, network, people and projects of entrepreneurs I work with 
not to have predetermined positions, and to be open minded about 
innovation and what I report.

At this time:

1. I raised a few questions to know what 1NET was about and what 
people (who are experts or in charge in their field) really think .

2. I thank everyone for the responses I received. What I can report so far is:

* 1NET represents a technical and political internet governance 
debate where Govs, Technicians, multilateral organizations, and civil 
society contributors are trying to consolidate their positions.
* without any consideration for sciences, reality, users, sovereign 
uses. The same for DNS, IPv6/IPv4, networking architecture, etc.
* one does not know who to eventually trust and why. MSism means 
political and technical positions, not decisions by others, but by oneselve.

3. The only consensus (actually opposed by everyone's "it cannot 
work") is that one should be able to test everything. It is also not 
documented by any RFC or charter about using the internet as its own 
test-bed for experimenting what is technically and politically disputed.

MG

At 07:08 04/01/2014, David Conrad wrote:
>Michel,
>
>On Jan 3, 2014, at 7:45 PM, Michel Gauthier <mg at telepresse.com> wrote
> > Questions to everyone:
>
>I'll play along.
>
> > 1. is RFC 1034 correctly understood by ICANN?
>
>Sure (modulo the myriad updates to 1034 that have occurred since its 
>publication).
>
> > 2. which responsible experimentation has ICANN fostered in 12 years?
>
>None that I'm aware of off-hand.
>
> > 3. It seems there is a subtile difference between:
> >
> > 3.1. the concept of "alternate root" (singular) which is opposed 
> on mathematical grounds that everyone can understand (two versions 
> of the same hierarchy that can pollute one another). This is what 
> you, Vittorio and Nathalie are discussing. I am not interested in 
> this ICANN/status-quo smokescreen issue we all know the no-interest 
> cons/pros by heart.
>
>I suspect you don't fully understand the "ICANN/status-quo 
>smokescreen issue" or you probably wouldn't reference it that way.
>
> > 3.2. the DNS architecture conceived to support a multiplicity of 
> fully separated "alternate roots" (plural) each under its separate 
> adminstration, rules, rates, purposes, AoC, etc. As there are 
> thousands private ones.
>
>Yes, the DNS architecture supports multiple classes. The 
>implementation of the DNS on the Internet today, and in particular, 
>_all_ the applications currently on the Internet, does not.  Talking 
>about using different classes to support "alternate roots" (assuming 
>that's what you're referencing) is a waste of time: instead of 
>updating every name server, resolution library, and application on 
>the Internet to support classes, it would be infinitely 
>easier/better to simply replace the DNS with something less wart-ful.
>
> > ICANN and IETF should clarify this terminology before Sao Paulo 
> as there are several "alternate root" (singular) administrators 
> (plural). and quite a few of private/public "alternate roots"A 
> press release would be enough.
>
>To be honest, I'm not sure what clarification you're looking for.
>
> > 3.3. what is the situation of the Chinese DNS?
>
>Since the delegation of IDNs, I'm actually unsure these days.
>
> > How the i-DNs plug-in qualifies?
>
>Depends on implementation I suppose.  If the plug-in partitions the 
>namespace to ensure their are no collisions, it has re-implemented a 
>singular root via however the partitions are imposed.  If it does 
>not protect against collisions, then would be fit into your 3.1 
>category above that you say you're not interested in.
>
> > As an alternate-version of the ICANN/NTIA class, as an alternate 
> root among the few roots, or as an alternate DNS?
>
>For clarification, there is no "ICANN/NTIA" class. There is class 
>"IN" (0x0001) used in the Internet, class "CH" (0x0003) formerly 
>used for MIT's ChaosNet, and class "HS" (0x0004) formerly used for 
>MIT's Hesiod service.  Presumably, if the plug-ins are supposed to 
>work on the Internet, they'll be using class IN albeit potentially 
>with collisions.
>
> > 4. ICANN has sold the exclusive uses of Internet TLD names 
> without specifying it was only for the "ICANN/NTIA" ("IN") class.
>
>If you say so -- I haven't bothered to look.
>
> > What does prevent anyone to set-up a "private-use class" global 
> name space, supporting the same and more or less TLDs as/than those 
> of the ICANN/NTIA class, that anyone using a "client or 
> applications software" also supporting that [private-use class] 
> namespace" may resolve?
>
>Absolutely nothing.  Of course, if they wanted to avoid collisions 
>with other folks using non-IN classes, they'd need to obtain an 
>allocation from IANA via "IETF Review" (see RFC 5226, section 4.1), 
>moving the singular root from the DNS into the IANA class 
>registry.  If they didn't care about collisions, then we get back to 
>your 3.1 category above that you say you're not interested in.
>
> > Should not inter-root administrators governance to be set-up to 
> avoid confusions, discuss IP issues, and foster coopetition be part of the IG?
>
>Before we pursue creating a superstructure to facilitate inter-root 
>administration, can you point to any non-trivial deployment of 
>server software, resolution libraries, or applications that support 
>a class other than IN?  If so, can you point to any non-trivial user 
>community with which discussions can occur?
>
> > It seems that this has a technical governance part, from what you say?
>
>Not really.  For there to be governance, I suspect there actually 
>has to be something other than a few lab experiments to govern.
>
> > 5. There are on-going rumor about the discussion in Sao Paulo of 
> specialized (experimental) classes, in particular in the "IoT" area 
> where Fadi Chehade would like to strike anIANA deal with GS1, as it 
> is permitted by the one you explained you signed for the IETF.
>
>If you say so. I'd be a bit surprised that the IoT folks would want 
>to retard their development/deployment cycles, but I suppose odder 
>things have happened.  However that's not a question.
>
> > It is important to realize that everyone want to know who is 
> leading the show and to where. There is a meeting that has been 
> agreed between ICANN and Brazil. It seems its preparation boils 
> down to an 1NET/LOG cooking. LOG does not respond. ICANN and 
> I*society disagree.
>
>It isn't clear to me things have progressed to a point where there 
>is actual disagreement.
>
> > Not easy to understand how corporate users and entrepreneurs 
> should consider all this, and what they should possibly participate 
> or alternatively proceed.
>
>True. However, I suspect the vast majority of corporate users and 
>entrepreneurs won't care until they see a potential impact in terms 
>of {dollars,yen,euros,bhat,etc}.
>
> > 5. So the last question for the day is : is there someone who 
> knows where all this is leading to?
>
>Um. Sao Paulo, Brazil? If you meant what is all this leading to, I 
>suspect more "fine lunches and dinners" in exotic locations :).
>
>More seriously, I believe it is an honest effort on the part of many 
>folks to provide an umbrella for multi-stakeholder "Internet 
>Governance" discussions as I suspect there are very few people who 
>understand the issues who think the status quo is ideal.  It seems 
>clear there needs to be some evolution in the structures/mechanisms 
>of "Internet Governance" (however you want to define it) and I 
>personally believe that the folks behind 1net.org are trying to 
>provide a venue where the issues can be discussed.  As a result, 
>logically, they're the target of the "Daily Hate". I imagine where 
>this actually leads depends on the willingness of the "Internet 
>community" (broadly speaking) to play along.  To be honest, it isn't 
>clear to me that the "Internet community" (broadly speaking) has 
>noticed there's an issue that can be addressed via venues like 
>1net.org, but I could be wrong.
>
> > Is there anyone in the Internet Cockpit?
>
>http://en.wikipedia.org/wiki/Decentralization
>
>Regards,
>-drc




More information about the discuss mailing list