Is it not the case that if you distribute an unique namespace (rather
than use a tree for DNS) you will end up swapping a root based DNS
architecture for some form of PKI to authenticate the distributed
namespace as meeting policy and that this also needs a structure to
guarantee authenticity and to achieve this universally we would end
up with some similar looking policy control issues to determine how
to manage the infrastructure so that it is safe?
Of course having decided on the operational policy parameters it
would be useful to be able to automate the operations. But surely the
same could apply to IANA functions? It's the policy that is tough and
takes thought.
Incidentally I agree that there is need of identifiers that users can
deploy that goes into the data infrastructure rather than simply the
underlying device (or pseudo device) infrastructure addressed by DNS.
But generally ideas along these lines that I've seen tend to
piggyback with or around the DNS rather than replace it. -
Christian de Larrinaga
cdel@xxxxxxxxxxxxx
On 30 Sep 2005, at 22:15, Michael Mealling wrote:
Steven M. Bellovin wrote:
In message <433DA1BD.1060707@xxxxxxxxxxxxxxxx>, Michael Mealling
writes:
Steven M. Bellovin wrote:
Reexamine the premises....
I am -- these are my premises. I lived far too long in the uucp
world to enjoy non-unique names; they led to nothing but trouble.
Again you're talking about mail routing and addressing mechanisms
when the people that use DNS in their web browser are looking for a
smart search interface that understands better what they're after
and why. Why do those two applications have to use the same
addressing scheme? Many of the political problems with DNS have
nothing to do with routing email and have everything to do with the
fact that its what your grandmother is using as an interface.
Some of the other requirements are security requirements, and
that's what I do for a living.
Sure.... security requires a level of exactness that you shouldn't
burden the user with or else he won't use the system....
I agree that the current DNS has serious problems, most notably in
the trademark sphere. That doesn't mean that its other premises
are wrong; there are other navigational systems that yield unique
results besides treees.
And what I'm suggesting is that uniqueness is a requirement of
networks and system, not people. The issues the UN has with the way
DNS is run have to do with the fact that you're trying to apply a
requirement of the network to society and that creates problems.
IMHO, we should look at building a system that works the way people
use identifiers and identity and then get that to work with the
existing network we have.
-MM
--
Michael Mealling Masten Space Systems, Inc.
VP Business Development 473 Sapena Ct.
Office: +1-678-581-9656 Suite 23
Cell: +1-678-640-6884 Santa Clara, CA 95054
http://masten-space.com/
_______________________________________________
Ietf@xxxxxxxx
https://www1.ietf.org/mailman/listinfo/ietf
_______________________________________________
Ietf@xxxxxxxx
https://www1.ietf.org/mailman/listinfo/ietf