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

[Phil Regnauld]

Karl Auerbach (karl) writes:
> 
>  Let's go back to my premise - consistency.  Competing DNS roots that are 
>  consistent will lead to minimal degrees of user surprise.  Are you surprised 
>  when you go into a super market and it has some extra boutique products on 
>  its shelves in addition to all of the standard, expected brands?

	Supermarkets and global, distributed information systems are not comparable.
	Not everything can be brought back to free market analogies.

> No, you 
>  aren't.  So why should users be surprised when they find that through a 
>  competing root they get the familiar .com, .net, .org, .arpa, .biz, .info 
>  and the 250 ccTLDs plus a few new boutique TLDs?

	It makes it harder for the user, not easier.

>  Is it "snake oil" if I hand someone a business card and say call me - and 
>  the telephone number on that card is "sip://1234@cavebear.com"?

	No it's a standard, but just like the fax, it only was *useful* when
	many other people had a fax.

	In the same fashion, I didn't hand out my email address to people
	around me in 1992 -- most didn't know what email was.

>  Which raises the question - are we inventing yet another consumer protection 
>  body?

	No.

>  If so, why?  And under what authority?

	That's your premise.

>  Moreover there are more than enough laws on the books about 
>  misrepresentation and fraud.

	Yes.

>  And do we want to assume that buyers are increasingly stupid and 
>  non-informed?  What about the intelligent and knowledgeable person who wants 
>  to, to usurp a marketing phrase, "think different"?  Do we deny that person 
>  in order to protect the rest?

	Feel free to create your own TLD locally, no one can keep you from
	doing that.

>  The logic being expressed is the logic that would deny IPv6 - for the most 
>  part people can not use IPv6 beyond their local nets.  So, should IPv6 be 
>  banned as "snake oil"?

	See my example above.

>  And perhaps we even ought to ban 4 letter TLDs because so much of the 
>  Javascript in web page forms around the world refuses to accept user 
>  contacts with 4+ letter TLDs in their email.  (Just wait until 
>  internationalized, IDN, TLDs come along - perhaps we ought to ban those too 
>  because they will also, in the eyes of some, be "snake oil.")

	Find me one website that accepts all valid RFC2822 addresses.  It doesn't
	exist.  It's besides the point.

>  On top of this there is no technical way to deny the rise of competing 
>  roots.

	There's a long way between not denying them and promoting them as a good
	thing.  Changes may be needed, but solutions looking for customers,
	we have plenty of.

>  One might ask "why in the world would one want to establish a competing 
>  root" (apart from the obvious answer that it escapes the USA hegemony over 
>  the current DNS)?
> 
>  One reason is that it can be a money making proposition.  It is possible to 
>  derive a very valuable stream of marketing data from the query stream that 
>  hits root and TLD servers.   In fact, Verisign has express permission from 
>  ICANN to do this.

	Oh, and alternative root outfits are doing this for the benefit
	of humanity ?

>  (And one can guess that those root servers operated by the US military and 
>  US government agencies are not quietly ignoring all the potential 
>  intelligence data that could be derived by watching the queries [and perhaps 
>  manipulating the responses.])

	Yes, I've heard of this.  Remind how many of these servers
	(the anycasted total, not which of A - L) are outside US territory ?

	No doubt the military are doing stats.  So would I :)

>  A prospective operator of a competing root system might induce people (or 
>  more likely their providers) to switch to their service by paying people to 
>  use it.  Imagine if you (or your ISP) were to get a check for $100 (the same 
>  unit of payment used in Google's AdSense program) every time you (or the 
>  ISP) resolved a million names?

	It might be good business model indeed, just like spam.

>  I mention Google AdSense because it is a good model - Just as Google pays 
>  web site operators to post Google provided advertising (for which Google is 
>  paid by the advertisers), a competing root server operator could attract DNS 
>  query traffic by peeling off a portion of the revenue from sales of 
>  marketing data derived from the query stream and paying that peeled-off part 
>  to those users that send traffic.

	Google does not manipulate people's traffic (yet).

>  I never cease to be amazed at how quickly people want to suppress the 
>  innovative and created spirit that created the internet in the first place.

	Re-read what I wrote earlier to Meryem: alternative technologies to
	continue to promote a unique namespace (single root) does not equate
	promoting alternate (multiple) roots.  And please don't try to infer
	that people who, while they might not be favorable to the current
	administrative model and to ICANN's disputed independence, don't
	see the necessity of having alternate roots, are trying to suppress
	the innovation and creative spirit that indeed led to the Internet.

>  When we started the net back in the 1970's - I was there - we were not able 
>  to interact with anybody else.  The common wisdom of that era was that data 
>  networks would be based on the then up and coming ISDN and that this packet 
>  switching stuff was ... well to use some words I've read recently, "snake 
>  oil".

	And now we're at a point where many, many people are depending on 
	the current system as it is implemented -- experiments are good,
	we don't have to be in the 70's do so, and finding new ways to
	explore unique namespace is one of them.  But fragmenting the
	current namespace is not helping.

>  In the 1980's when I formed my first two internet based companies not many 
>  people could send email to "karl at epilogue.com" or "karl at empirical.com" - in 
>  those days "real" email was from MCI or IBM and others.  Internet email 
>  addresses were, again to use some recent words "snake oil".

	What's your point ?  Are you saying that email, which was open,
	replaced the arachic MCI Mail, Bitnet, and Compuserve ?

	Well, at first we had UUCP maps, which didn't scale that well,
	and guess what displaced _them_ ?

	An addressing mechanism based on a unique hierarchical namespace.

	Shall we go back to Fidonet ?

> > 	How am I going to validate ".boutique-tld" if my nameservers don't
> > 	know about it ?
> 
>  What means this word "validate"?  How does one today "validate" 
>  gdfkjljd.xn-r5tyk8dkjui.com?

	Validate: look it up, confirm it exists.

>  DNS is not a system of "validation".  Attempts to use it as one are like 
>  attempts to build balloons out of stones.
>  If you want more, then one needs to move to mutual identification and 
>  authentication mechanisms such as IPsec.


	That was not what I was talking about.

	Using DNS as a validation framework for other application may be
	awkward, but I was talking about testing the validity (as in: is
	it registered ?  does it have an MX ?) of the domain _itself_.

	
>  If you don't like TLDs not approved by <insert your favorite TLD creation 
>  body name here> then don't accept 'em.

	Indeed, that's easy.
	
> Gaining that acceptance is part of 
>  the gauntlet that a boutique TLD needs to run - at its own expense and 
>  through its own efforts.

	Good luck to them.

>  I see no reason to institutionalize any kind of help or assistance to any 
>  new TLD aspirant.  The costs to set up a new TLD are small.

	I agree with you on that.  Many more TLDs should easily be accomodated
	by the current model.

>  Take for example my .ewe TLD - 

	I would call it "side level domain", I can't find a delegation for it.

	What mechanism should I use to locate it ?

____________________________________________________________
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