Competition ? Re: [governance] is icann ...

Karl Auerbach karl at cavebear.com
Mon Sep 21 17:22:04 EDT 2009


On 09/21/2009 10:14 AM, Dr. Francis MUGUET wrote:

> except for some fields like the CNAME ...

I invite you to dig into the following domain name, which has a single 
CNAME resource record... maps-to-nonascii.rfc-test.net

As the name suggests, it uses a CNAME RR to load an indirect name that 
contains unusual, but 100% Internet-Standard legitimate characters, such 
as carriage-return, linefeed, and null.  That indirect name causes 
things like gethostbyname() on *nix systems to fail.

There are uncharted traps in the DNS world that will be exposed as we 
start to stretch DNS into parts that are legal but have not been 
exercised, parts such as using DNS classes other than IN.

I am in the business of testing internet code.  We test not so much for 
conformance but for robustness.  And we, unfortunately, have learned 
that much internet code is nowhere close to being worthy of being 
labeled as "robust" but is instead implemented according to a design 
practice that says "it worked once on our unstressed lab network, so 
let's ship it as a finished product".  That kind of code is often very 
brittle; it often breaks when subjected to uses beyond what the norm of 
today.  (And besides inducing failure, user confusion, and ISP support 
headaches, it opens doors for security penetrations.)

It is becoming increasingly clear that DNS is simply not adequate for 
the name mapping needs for the future.  We need name mapping systems 
that are more agile and have greater temporal stability than DNS, and 
they need to avoid the trademark entanglements that have strangled 
rational DNS policies.

What I am suggesting is that rather than trying to patch DNS and dealing 
with the mass of the installed base we should concentrate on better 
technolgies that perhaps map to an underlying DNS base, perhaps not, but 
that are themselves more attuned to the developing needs of the net.

Of course, these new technologies might need new bodies of governance. 
They would likely be new bodies, not ICANN, thus giving us a chance to 
do governance things well from the outset rather than do a retrofit.

		--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