[governance] RE: Human rights and new gTLDs

Karl Auerbach karl at cavebear.com
Wed Sep 26 14:29:20 EDT 2007


Vittorio Bertola wrote:

> This also applies to Karl's reply - Karl, I do respect your freedom to 
> live as you like, but when we move from the offline to the online world, 
> and speak about something at the root level of the Internet, the entire 
> planet is packed in the same room.

Singularity is not a technical requirement of DNS.

There is no need for a single DNS root.  For years already there have 
been many such roots, most ill run and laughable, but some are 
professionally managed by people who really know what they are doing.

First to answer McTim's question - I really don't understand the nuances 
of DNSSEC to know how it affects multi-rooted DNS.

Back to the notion that the internet has and must have one singular, 
authoritative catholic root:

That myth is the key to ICANN's authority, which is why ICANN so closely 
protects that myth.  Yet is a myth, a dogma of faith, nothing more.

If one watches how DNS names are resolved, I mean really look at the 
deep technical level, even to the degree of watching the "authoritative 
answer" bit, one observes that it is system of referals.  So it does not 
matter how a bit of resolving software reaches the name server that has 
the target resource records, what matters is that it reached that name 
server.  As such, as long as separate systems of roots cause a resolver 
to reach the same name servers then the users get the same answers.

The issue here is consistency of DNS answers, not singularity of DNS roots.

In other words, different communities can shape their DNS landscapes 
differently to meet their own values and needs.

Consistency does not mean some mindless uniformity across the entire 
world - that road would equally require us all to speak the same language.

There are those who fear that a DNS name contained in a URL will somehow 
resolve to two different things for two different users.

That is a misplaced fear: DNS can never guarantee consistent results, 
even with a single root.

In other words the thing that is so feared and is used as an argument 
against competing systems of DNS roots is a thing that is intrinsic to 
DNS itself and exists even with a singular DNS root.

I wrote a note on the chimera of DNS as a global uniform internet name 
(GUIN) space - 
http://www.cavebear.com/archive/rw/nrc_presentation_july_11_2001.ppt.

I found three kinds of properties that are required for a global 
internet name:

	- Universal validity or non-validity (i.e. every name that is valid 
must be valid everywhere and every name that is invalid must be invalid 
everywhere.)

	- Location invariance: every name must have the same meaning no matter 
where uttered.

	- Client invariance: every name must have the same meeting no matter 
who utters it.

	- Temporal invariance: Once a name obtains a meaning it must retain 
that same meaning for all future time.

DNS does not possess these properties.

Content management systems and optimizers have made short work of 
location and client invariance.  Filters have made mincemeat of 
universal validity or non validity.

But more importantly DNS does not possess temporal invariance - DNS 
names change over time, often rather quickly - and as such they are 
inappropriate vessels into which to put our hopes for names with 
unchanging meaning.

All of us daily experience DNS failure of temporal invariance - we have 
all observed how email addresses and web URLs rot into meaningless or 
become handles to new, unexpected, targets.

Given that DNS is already a week vehicle the goal of a single catholic 
root becomes rather less important because even that singular vehicle 
can not obtain the desired outcome of universal names.

I have come to believe that we should view the DNS system as one in 
which there can be separate roots, each with its chosen suite of TLDs.

Rational self interest will drive the operators of these roots to 
include in their inventory the core TLDs with which we are all familiar.

However, local, aspiring, boutique TLDs will be found in some roots, not 
in others.  This is where the competition for air and light occurs; this 
is where a new idea grows.  And this is where communities will prune and 
shape their view of the internet landscape according to their own 
desires, values, and aspirations.

As those boutique TLDs grow, or fail, they will be adopted, or not, by 
more root systems.  And name collisions can be resolved among the 
competing claimants using the traditional national and international 
legal mechanism and systems used to resolve among multiple claimants to 
a trade or service mark.

Users can pick and chose which root system to which they will subscribe. 
  If they don't like the offerings of root A, they can move to root B, 
just as today if they don't find their favorite boutique brand of soup 
at supermarket A they can go to supermarket B.

It is for this reason - the reason that DNS can operate with multiple 
roots, each with its own suite of TLDs chosen by the operator - that I 
do not consider DNS to be a critical internet resource that has to be 
managed by an overlord of names.

But that simple idea, like the idea of packet switching in the 1970s, 
scares established institutions.  And as AT&T tried to bury the 
Hush-a-Phone under techno-FED, established institutions try to cast the 
fact that DNS technology, without change, can have multiple roots, as 
some kind of anathema.

		--karl--
____________________________________________________________
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