[governance] [discuss] [bestbits] The decentralization of the DNS system

McTim dogwallah at gmail.com
Sun Jun 21 09:22:53 EDT 2015

<cc list  trimmed to lists I am subbed to>

On Sun, Jun 21, 2015 at 5:23 AM, Jean-Christophe NOTHIAS I The Global
Journal <jc.nothias at theglobaljournal.net> wrote:
> Adebunmi, Willi,
> I'd like to try to explain why the "decentralization" idea is one of the
> very misguiding conception in the field of Internet Governance, as I feel
> this is a very "core" part of the overall misunderstanding

Agreed, there is a misunderstanding.

> When McTim, and Norbert respectively mention their view as:
> McTim: "The DNS is already decentralized by it's very nature. I think the
> word the Willi seeks is 'non-hierachical'".
> Norbert: " The DNS has been designed as a highly decentralized system with a
> very lightweight root"
> In the Internet, what is truly  "decentralized" are the networks
> (infrastructure level). They belong to separate entities. They are managed
> by different authorities whether private or public or public/private. They
> are by nature  decentralized by opposition to a centralized network. These
> networks are interconnected by technical means, creating a virtual single
> space where packets can circulate from one gate to another according to
> protocols, from one IP address to another IP address.
> In the Internet, what is truly "centralized" is the management of the DNS
> (governance issue).

No, the correct term for this is "hierarchical.  The DNS is
DE-centralised.  Nameservers are run by local entities in much the
same way that routers are run by local entities.  In theory, you
control your own router, you also control your own DNS server.  In
practice, most users leave those chores to service providers.

>  We all understand what, by extension, the notion of
> "very nature" means. It means by conception, or to be even more precise by
> human conception of an information technology (IT) system. So indeed nothing
> "natural", but an artificial IT engineering conception. Nothing wrong with
> that, but just to highlight the fact that there is no "natural order" in the
> Internet conception. Only intelligence by computer scientists and telecom
> engineers.
> Why is it possible to claim that the DNS management was conceptualized as
> being highly "centralized" (and not "decentralized"). Simply because in
> order to have one "directory boutique" that can reply correctly (i.e. link
> to files located at an IP address) when asked about connecting a visitor to
> a domain name, you make it easy only if this "directory of
> name-to-IPaddress" has every single name at hand, and is able to allocate
> domain name with no duplication of ownership, is able to centralize all
> updates/changes in the directory. In that sense, one can write that "by its
> very nature", the DNS management needs to be highly centralized. Once it is
> very centralized it is essential that it will be highly "distributed" with

In order to have the "coherence" described above, you need hierarchy,
not centralization.

> Now looking at the governance of the Internet, one core element of the
> dominant/asymmetric narrative, beyond the idea that it was an open bar
> (multistakeholders drinking beer on an equal footing), is that precisely
> because the DNS management was by its "very nature decentralized", so was
> its governance, leading to the fairytale that "no one controls the
> Internet".

There is no ONE entity that controls the Internet.  No one can
possibly do this as the "Internet" isn't a "thing" that can be
controlled per se.  It is a network of networks, each controlled by
different entities.

 CQFD. We can understand why this sounds smart to qualify or label
> the Internet Governance a decentralized space.
> The DNS management has to be somehow "hierarchical" based on the old
> master/slave IT concept - that is still the best way to make sure that we
> are all using only one name-to-address directory.
> Now, coming to an even more interesting point.
> Why should there be a single "directory" when it is possible to have plenty?


> Let's keep in mind that the word directory is wrong in this context. A
> directory is not supposed to allocate domain name to anyone, or to allocate
> TLDs whether ccTLDs or gTLDs to anyone. A directory has a basic database
> management role, highly centralized and distributed thanks to its many
> slaves.  Now a boutique like ICANN is doing much more than a directory.
> Let's suppose that we are familiar with all what ICANN does, still, it is
> very possible to have several ICANNs. One of them is OPEN-ROOT. It has a
> different philosophy - as an example, you pay your TLD once and for life,
> still it has the ability to reply to anyone with a request related to domain
> names sold by ICANN and its slaves affiliates (distributed network of a
> highly centralized function). Moreover Open-Root brings in new TLDs, and
> allow people to use more of these, safely. Moreover, there should be even
> more OPEN-ROOTS; that is no threat to the so-called OPEN-FREE-DECENTRALIZED
> network of networks (the term decentralized s correctly used here). By doing
> so, it introduces competition in a sector highly monopolistic. So each of
> the users would benefit from that competition, not endangering the fair use
> of Internet. Again, I am actually using Open-Root to browse the web, and I
> am myself a registry/registar for the gTLD .ngo.

So you are a Registry operator for the TLD .ngo.  So is PIR.  If you
sell mctim.ngo to me, but someone else registers mctim.ngo via PIR,
then coherence is broken.

 Interestingly there is no
> real need for these intermediate function with Open root. I did this by
> paying 200 € paid to Open-Root. And to be fair, let me correct McTim by
> saying that Open-Root offers more than ICANN as Open-Root edits new TLDs
> (not seen by ICANN actually)

yes, they offer a broken Internet.

> McTim writes very rightly that:
> "If a TLF is not published in the IANA root, the vast majority of users
> can't see it".
> This is a critical issue. Because users are not yet aware that they are
> alternatives, and because all computers  (distributed central power), the users are again part of capture
> audience with no choice. But it doesn't mean that over time this
> domination/capturing will sustain.

Demonstrably false.  I just bought a new computer and there were no
"DNS settings are pre-set to an ICANN slave".

> Would civil society representatives (whatever legitimacy they claim they
> have) call for a competitive approach to DNS management, and not just
> basically support transition of IANA from ICANN to ICANN, would they support
> at least one single alternative root zoot management operator, they would
> gain leverage in terms of influencing and changing the unbearable asymmetry.

probably not.


> This competition would immediately pave the way to solutions that might fit
> to Willi's hope and demand.
> The power of commercial digital players, the advantages they gain from this
> single DNS rootzone management is huge and not acceptable if anyone has in
> mind an idea of public interest, or even considers the Internet as some sort
> of global common good. Obama said something like this lately.
> Having a competitor would help to self regulate the Internet, without having
> to build-up digital Ligne Maginot or nation state totalitarian system.
> But a few benefit greatly from the current state of things, and are
> obviously not willing to give away their privilèges.
> Old story
> JC
> McTim. Thanks for your comments. Let's have Louis adding his own comments,
> and more specifically what are his views of the surveillance aspect of
> alternative root zone management. Rootzone management has little to do with
> routers by the way.
> Le 20 juin 2015 à 08:10, Adebunmi AKINBO a écrit :
> Jean,
> Much respected response.
> Thanks for the consideration.
> Regards.
> -Akinbo.
> On Sat, Jun 20, 2015 at 6:47 AM, Jean-Christophe NOTHIAS I The Global
> Journal <jc.nothias at theglobaljournal.net> wrote:
>> You are very right Adebunni. Sorry for my advise. I just didn't want to
>> bother anyone with simple ideas that work, and that will soon or later
>> transform the current asymmetry.
>> JC
>> (the lists are not always the best venue to "share" as we have to face a
>> lot of negativity in here)
>> Le 20 juin 2015 à 07:40, Adebunmi AKINBO a écrit :
>> Willi,
>> Whatever you do, do not take Jean's advice to write directly to him.
>> You two need to share your opinion with people like me or the world.
>> How best does Africa begin to learn and participate without you both
>> sharing your thoughts?
>> I do appreciate it. Both of you.
>> Regards.
>> On Sat, Jun 20, 2015 at 6:02 AM, Jean-Christophe NOTHIAS I The Global
>> Journal <jc.nothias at theglobaljournal.net> wrote:
>>> Willi,
>>> Thanks for sharing your thoughts. If I may put two comments on this.
>>> 1_
>>> In my opinion, "decentralization" seems not to be the appropriate word to
>>> describe what and how to change the current monopole under ICANN.
>>> Information Technology is somehow always related to a Master and its slaves,
>>> by electronic nature.
>>> 2_
>>> Localisation is often associated with the idea of "nation". Keep in mind
>>> that this could mean to imprison people into old boundaries. Localisation
>>> might be interesting if a community decides to set up its own network (see
>>> the Spanish experiment on this) but that does not address the DNS issue.
>>> In other words, decentralization has been a buzz word propagated by the
>>> current owners/rulers of the DNS root zoot management. And basically it is
>>> part of the dominant narrative related to the so-called, open, free,
>>> decentralized Internet under US/allies ruling boot. Localisation might
>>> equate to a returning in the past, pushing us back within the boundaries of
>>> the old national thinking. Not sure if we really want this.
>>> What is more needed is either a global common governance (option one),
>>> with a public interest perspective, or a competitive market. Were we not
>>> satisfied with the ICANN, we should turn to another root-zone manager. This
>>> is no dream or utopia. I am no longer sending my domain name request to an
>>> ICANN affiliate server, but instead using the Open-Root system to find
>>> whatever I am looking for on the web. Thanks to Open-Root, we are also
>>> providing for free one domain name with a gTLD managed by Open-Root to NGOs.
>>> When the new gTLD .ngo by PIR (Public INterest Registry) given by ICANN to
>>> ISOC (PIR is ISOC's TLD roommate and milk cow) is an additional business
>>> supposed to make more money, we are happy to provide access to IPs through
>>> an independent, cheap (for free, or paid for life) domain name. All our
>>> computers are using Open-Root DNS management to access website that ICANN et
>>> al cannot see if we do not want the US surveillance apparatus to see it.
>>> The first option (Global Common Governance) is almost dead, thanks to the
>>> systematic blockade by the US (gov and businesses) and its usual allies.
>>> Moreover, this first option would require both an architectural re-thinking
>>> (see JFC's email) and a political and institutional framing (see JNC for its
>>> democratic approach of the Internet governance). A long way to go. You show
>>> note that the request for a roadmap to a new Internet Governance, as put
>>> before the Net Mundial Conference has gone no where expect into giving to
>>> ICANN more power over the IANA functions (shifting power from the US to the
>>> US).
>>> The second option is fair competition (which I like as it means ending
>>> the de facto ICANN monopole) and we are free to practice competition it at
>>> any time starting today.
>>> A third option is an old fashion scheme that would fragment the Internet
>>> into national sub-Internets, (Westphalian Internets). This is not just
>>> old-fashion. This would be a way to imprison people back into their country
>>> land under the control of their leaders (good luck with that), unless the
>>> current efforts by a few academics come to conclusion in order to
>>> interconnect different root-zone management systems. There are a few bright
>>> minds working on this interconnectivity, whether the roots would be national
>>> or global.
>>> For anyone interested to use the OPEN ROOT to browse the web, and break
>>> free from the ICANN affiliates, feel free to write to me for guidance and
>>> information.
>>> JC
>>> Le 19 juin 2015 à 23:28, willi uebelherr a écrit :
>>> The decentralization of the DNS system
>>> We need a completely self-organizing Internet. And this is possible only
>>> through massive decentralization. We can look at the difficulties at the
>>> beginning of the Internet with tolerance. They were mostly of technical
>>> nature. But today we have other conditions. And under such other conditions
>>> arise other possibilities.
>>> It is about the IP address. It is necessary to ensure that the packets
>>> find their direct path to their goals. The router work with numbers. We
>>> humans with text. The content is the same. Only the representation is
>>> different.
>>> The Internet, a transport system for digital data in packet form, needs
>>> the destination address in order to direct the packets to their
>>> destinations. The packages contain that destination address. Thus, the
>>> packets are always the instance to activate and orient the router.
>>> This, however, requires the knowledge of the geographical position of the
>>> target in order to determine the direction to this. But this question is not
>>> the subject of this text. Here i speak about about how the transformation of
>>> a text can be organized properly into the numerical representation of the IP
>>> address.
>>> We focus on the ccTLD (country code Top Level Domain). It is the first
>>> level of the cascade for the decentralized organization of the DNS system.
>>> The gTLD (generic TLD such as .com, .org, .net ...) do not interest us. Each
>>> person can decide for themselves whether they want to apply this nonsense.
>>> Each host on the Internet, client or server or both, has a unique and
>>> singular geographical position. We describe it with the world coordinate
>>> system WCS 1984 (WCS84) or later versions. All GPS and online map systems
>>> work with that. For mobile devices, this is always the position of the
>>> access point to the Internet. To transport the packages we need this
>>> information so that the router can select the most direct route.
>>> The decentralization of the DNS system rests on the cascade steadily
>>> reduced regions. The first level is the ccTLD. So a country with borders, as
>>> we know it today. We can use this, although it is not optimal. All other
>>> levels are determined exclusively in their regions. The administration, as
>>> set, change, and resolve, only happens in the region. This applies to each
>>> level.
>>> The organizational level for the ccTLD is the association of the
>>> countries on our planet. So the UN. The regions within a country are
>>> organizational objects in the respective country. Local regions are
>>> organizational properties of the larger region.
>>> The organizational models are always determined at the level of each
>>> region. In order for a region in Brazil is achieved via the ccTLD ".br"
>>> worldwide. But their internal deeper structure may be different from the
>>> structure in Kenya, India or Russia. Regardless of the specific local /
>>> regional organizational structure of the DNS system, we always get the
>>> correct IP address for our goal. Only the principles of the resolution of a
>>> domain sequence into a numeric IP address is the subject of our common
>>> discussion.
>>> With the local self-organization we dissolve the need for global Internet
>>> Governance. This may for organizations, that are derive its raison d'être
>>> from the global Internet Governance, be uncomfortable. We carry this with
>>> serenity.
>>> Important for us is to help all the people in the different regions of
>>> our planet to organize their own DNA structure in accordance with their own
>>> principles. This makes it possible for all people of our planet to connect
>>> to all regions of our planet for the communication.
>>> We have several instruments which have well proven in the history of the
>>> Internet. The most important instrument are the RFCs. With that we can best
>>> explain the principles of decentralized DNS system. And this is also the
>>> place where we describe the global access to the ccTLD's. The rest is
>>> regional and local task.
>>> The dynamics in the inventory of domains can be very large. But this task
>>> is clearly and simply by  decentralization of the administration.
>>> An important field of our activities in the IG forums is the propagation
>>> of a free access to the setting up, modification and dissolution of a
>>> domain. Technically this is not a major challenge, because the processing of
>>> an item can be organized by the applicants themselves. There is only one set
>>> in a simple database. The blockages are in the bureaucratic systems. But the
>>> dissolution of these blockades always remains the task of the people in
>>> their regions.
>>> With the help of free software and open source software we can do this
>>> very easily realized in a large cooperation. So the DNS system is an
>>> experiential field of creative and international cooperation.
>>> with many greetings, willi
>>> Porto Alegre, Brasil
>>> ____________________________________________________________
>>> You received this message as a subscriber on the list:
>>>     bestbits at lists.bestbits.net.
>>> To unsubscribe or change your settings, visit:
>>>     http://lists.bestbits.net/wws/info/bestbits
>>> _______________________________________________
>>> discuss mailing list
>>> discuss at 1net.org
>>> http://1net-mail.1net.org/mailman/listinfo/discuss
> ____________________________________________________________
> You received this message as a subscriber on the list:
>      governance at lists.igcaucus.org
> To be removed from the list, visit:
>      http://www.igcaucus.org/unsubscribing
> For all other list information and functions, see:
>      http://lists.igcaucus.org/info/governance
> To edit your profile and to find the IGC's charter, see:
>      http://www.igcaucus.org/
> Translate this email: http://translate.google.com/translate_t


"A name indicates what we seek. An address indicates where it is. A
route indicates how we get there."  Jon Postel

-------------- next part --------------
You received this message as a subscriber on the list:
     governance at lists.igcaucus.org
To be removed from the list, visit:

For all other list information and functions, see:
To edit your profile and to find the IGC's charter, see:

Translate this email: http://translate.google.com/translate_t

More information about the Governance mailing list