Alternative DNS systems and net neutrality - Was: Re: [governance] DNSsec and allternative DNS system

Meryem Marzouki marzouki at ras.eu.org
Thu Nov 15 13:14:22 EST 2007


Le 15 nov. 07 à 12:28, Suresh Ramasubramanian a écrit :

> Meryem Marzouki wrote:
>
>> Then please tell me where it's happening at global level, with all
>> interested parties involved and with prior agreement on the principle
>> that such global public policy is desirable. For sure, it's not at
>> IGF. And it's not even on this list. Unless I missed something.
>
> I wish you luck
>
> * Finding organizations that can offer viable and well managed  
> alternate
> roots / TLDs

Unless there is a misunderstanding on the "alternate to what"  
meaning, you're not really saying that this is an issue, are you?  
Imagine simply, in the current situation, if the current 15 gTLD were  
under, let's say, 3 different roots, and all current ccTLDs were  
grouped into, e.g. 5 additional roots, or even just one. Don't you  
think that these 8 roots would be well managed and viable ? Don't you  
think this could also be the case for additional roots (it is,  
already)? If there are currently registries able to operate TLDs, why  
wouldn't there be more organizations able to operate roots?
Let's talk about coordinated distinct roots, if "alternate" or  
"competing" may lead to misunderstandings.

> * And that can actually achieve consensus on the mechanisms these are
> published with

That's the main, tougher issue, and it's a global governance issue.  
But there are no chance to achieve consensus if we don't start  
drafting the possible options, or even start thinking we could have  
such a discussion..

Let's suppose that we engage in such a discussion. What should be the  
rules to achieve consensus on? I see three sets of rules:

- One set of technical rules, that form the basis of a technical (and  
only technical) commitment (or MoU, if anyone prefers). We need to  
identify the set of #T1 to #Tn technical conditions  such that, if  
not met, may endanger the operation of the whole system. Honestly, I  
don't know how they can translate, I'm not competent to enter this  
discussion, but these conditions are already well known by relevant  
people.

- One set of "behavioral" rules that form the "behavioral" MoU. Let's  
start drafting some of them:
Rule #B1: each TLD should be unique, whatever root it is operating  
under. This is the sine qua non condition to have the whole system  
working. (NB. yes, it's a behavioral rule, not a technical rule)
Rule #B2: any TLD under any root in the system should be resolved by  
all roots part of the coordinated system.
Rule #B3: all roots part of the coordinated system should stop  
resolving TLDs under a root that has been banned from the coordinated  
system
Rule #B4: any root part of the coordinated system may stop resolving  
TLDs under a root against which x% of all roots has voted
etc. I'm not sure these rules are good or bad or even necessary  
(apart from #B1 and #B2). My point is to show what such rules could  
look like.

- One set of global cooperation rules that form the global  
cooperation MoU:
Rule #C1: the coordinated system is open to any new root, provided  
that it signs the technical and the behavioral MoUs
Rule #C2: defines the decision making system (one root one vote? or  
any other to be discussed)
Rule #C2: a root infringing any of the behavioral rules is  
automatically banned from the system
Rule #C3: banning a root from the system for any other reason than a  
behavioral rule infringement requires unanimous decision
Rule #C4: a root infringing any of the technical rules should be let  
aside of the system until things are fixed.
etc. same disclaimer as above for the "behavioral" rules.

Many, many, many issues are still unresolved. Like is there a rule  
favoring or not preemption of TLDs (if you see what I mean:)) or is  
it first come first served basis? How could we avoid (policy, not  
technical) deadlocks? What is a root, i.e. we shouldn't end with one  
TLD == one root! And, the mother of all issues: how many IGFs would  
it take to have such a discussion on the table..

> Tough? Yes.

Yes, indeed!

> Stick to the existing root server model? Definitely, yes.

Not necessarily.

Best,
Meryem____________________________________________________________
You received this message as a subscriber on the list:
     governance at lists.cpsr.org
To be removed from the list, send any message to:
     governance-unsubscribe at lists.cpsr.org

For all list information and functions, see:
     http://lists.cpsr.org/lists/info/governance



More information about the Governance mailing list