Berkman
 Center for Internet and Society
The Debate Over Internet Governance:
A Snapshot in the Year 2000
 

 

Introduction


Interviews

    Karl Auerbach
    Fred Baker

    John Perry Barlow
    Dave Crocker
    Jay Fenello
    Carl Kaplan
 
    Michael Krieger
    Jamie Love
    Eric Menge
    Charles Nesson

    Mike Roberts
    Joe Sims

    

Topics
  Consensus
   The Future
   Governance
   ICANN
   The Internet
   Participants' Internet
      Background
   Participants' Biographies


Links


Message Board



Contact Us

 

 


 


Karl Auerbach

I.                  BIOGRAPHY

II.               PERSONAL BACKGROUND IN INTERNET ISSUES

III.           GOVERNANCE

a.     Defining Governance

b.     Does the Internet Need Governance?

c.     Is ICANN Governance? 

d. Governance vs. Technical Management

e.     Let the Market Regulate the DNS

f.     Structuring Global Internet Governance

g.     Policy Questions vs. Technical Questions

h.     How Would You Govern the Internet?

i.     Approach

IV.             ICANN

a.     Alternatives to ICANN

b.     Where ICANN Went Wrong 

c.     Can ICANN be Fixed?

d.     Limits on ICANN’s Authority

e.       The Worst ICANN Can Do (Part I | Part II)

V.                CONSENSUS

a.     Defining Consensus

b.     Alternatives to Consensus

VI.             THE INTERNET

a.     How It Works

b.     Commercial Interests

c.      What is the Internet’s Greatest Promise?

VII.         THE FUTURE

a.     ICANN In 10 Years

Please note that this transcript has been edited and redacted according to the author’s request. 


February 28, 2000

GOVERNANCE: Defining Governance

Q:  How do you define governance?

A:  Before I begin, even though it says Cisco on my business card, I don’t speak for Cisco.  I always make that clear; these are my own opinions.  What’s the question again?

Q:  How would you define governance?

A:  Mandatory control.  Where you don’t have an alternative; you really can’t say no. 

GOVERNANCE: Does the Internet Need Governance?

Q:  Do you think governance is necessary for the internet?

A:  In part.  Governance can be applied by a number of mechanisms.  You can have mechanical governance, a lot like ICANN, where you have a body that sits there and thinks and applies … rules, regulations, laws, whatever you want to call them.  You can also have governance in the Adam Smith style: the invisible hand.  Governance by economic forces can be just as powerful, if not more powerful in many respects, than formal governmental bodies. 

One thing about governance.  People are really afraid of that word.  The Reagan legacy has left us with government as an anathema.  But there’s real good reasons to have governance.  And there are parts of the internet that need it.  There are certainly parts that need it much more than others.  IP address assignments, which is under ICANN, is very much in need of regulation because of the relationship between the technology and the assignment of addresses.  And the economic impact of IP address assignment is enormous, and most people don’t even have a clue how the impact of that is going to have on the growth and development of the internet.  But, as far as … the other part, the domain name space, under ICANN, I assert that there is no need whatsoever for any regulatory body over it, other than allowing the standard forces of laws that apply over trademark infringement to apply. 

GOVERNANCE: Governance vs. Technical Management

Q:  What do you think of the representation that handling IP addresses is just technical management and not really governance? 

A:  I have a definition of what constitutes “technical” and what constitutes “governance.”  And my definition is: if you can make a wrong decision about the thing and that causes the internet to fail in delivering its fundamental service, then it is a technical matter.  And I define the fundamental service of the internet as the delivery of IP packets from one end to the other.  So if one makes a mistake about assigning addresses, it has a reasonably direct impact on the ability of the internet to carry packets from one end to the other.  If one makes a mistake about assigning domain names, there’s practically no impact on whether the internet has the ability to deliver packets from one end to the other.  So, to me, what is technical are things that have a very immediate impact on the ability of the net to deliver its services.  The governance being that, or the policy-making which is that, which is not directly related.  They may both involve technical questions.  Now you asked … there was another part of your question.  Can you restate the question?

Q:  … Just what is your response to the argument that IP address assignment is just technical management or plumbing?

A:  … IP addresses sort of occupy a slightly … according to my test, they’ re in a fuzzy zone because if one makes a mistake with assigning IP addresses the effect is going to be not immediate, but it’s cumulative.  If we assign enough addresses incorrectly, the routers of the internet, the things that Cisco and Nortel and Lucent and 3Com build, will have routing tables so large that they will consume a lot of memory.  The protocols that carry the routing information around will be so burdened with carrying updates that they will simply clog the net.  I don’t know if you understand how internet routing works.

THE INTERNET: How It Works

Q:  Yes, basically, but can you explain it for the record?

A:  Basically, every packet carries a destination address and the routers get the address, they look at a big table, and they say how do I get to the destination?  Or, how do I get closer to the destination?  It’s a lot like if you were flying from here to my house in Santa Cruz, … you wouldn’t have in your little address book actually how to get from here to my house.  You would have how to get to San Francisco.  Then from San Francisco you would look up, well how do you get to Santa Cruz.  Then from Santa Cruz you would look up how to get to Carbenero Drive.  Just sort of a progressive revelation.  Well we do that with the internet; it’s called address aggregation.  The idea is we want to be able to take the addresses and clump them together according to the typology of the internet so we only have to send around routing information pertaining to where these clumps are.  So you only have to know how to get to California, you don’t have to know how to get to Carbenero Drive in California. 

If addresses are not assigned in a cohesive way that is in accord with the typology of the network, then we have addresses that would be scattered everywhere and we would literally have to have a separate routing entry for every single possible destination, and that would be very much untenable.  But because of the fact that we have addresses that have to be assigned in accordance to typology, we are pouring concrete around the typology of the network, and if you think about the typology of the network, it’s just a small number of backbone providers, some ISP’s that do local delivery, and then the little guys who deliver.  Well, these little guys no longer have the option of switching to another upstream provider because their addresses are locked in.  They may have to get another set of addresses and they’re going to have to tell their customers to re-assign addresses, which is a big pain.  So IP addresses have the technical part, which is if we do it badly, then the network will start to crumble in terms of just having too many addresses.  But it has a policy part, insofar as we are imposing a structural regime on the internet relating to the big guys feeding the next tier, feeding the next tier  And we’re removing the ability of end users and the intermediate providers to switch their long-distance companies without a great deal of thought and economic pain.  IP version six will make that easier, but that’s not here yet.  So IP address assignment is not just plumbing.  But it’s got a strong technical component, it’s got a strong non-technical component.  Things like DNS have very few technical components at all. 

GOVERNANCE: Is ICANN Governance? 

Q:  So is ICANN governance?

A:  ICANN is governance with a vengeance.  The worst form of governance.  Arbitrary, capricious, imposed without any input from those who have to pay the taxes and suffer its regulations.  It is an oligarchy.  It is a business-run oligarchy.  It is a secret society.  Do I support ICANN?  I support the concept.  Do I support ICANN as it is?  No, I think it should be dismembered, right down to the ground. 

ICANN: Alternatives to ICANN

Q:  And what would you put in its place?

A:  I would put a distinct body to handle IP address assignments and I’d leave the DNS completely alone. 

GOVERNANCE: Let the Market Regulate the DNS

Q:  Why do you think the DNS has been put under ICANN’s regulation, since you seem to feel strongly that that’s not its place?

A:  Well there’s this premise – this is part of my talk [at Boston University] tomorrow [about] what is the internet and how do we understand what it is – there’s this notion that people believe that the domain name system is a fundamental service of the internet.  Is your telephone book a fundamental service of your telephone system?  25, 30 years ago people said, yeah, if you get from Ma Bell, you get the official statement of names and addresses.  No, but today you can get a phone book from all the local phone companies, you can get it from third parties, you can get it on CD-Rom, you can go to the web and look up phone numbers.  Does it matter?  No, as long as you get the number you want to call.  It works.  And if you get the wrong number, you just don’t use that service.  The DNS is just a service.  It’s just a service through which you present a character string saying, you know [e.g.] www.cavebear.com, and you say look it up for me.  It’s like I take my clothes in to be drycleaned.  I could have lots of drycleaners do the service.  I could have lots of DNS services do it.  There’s no necessity in the internet for having one DNS. 

Now there’s some technical flaws in the design of DNS that cause … People can build what are called zone files, and which define – like mine is on kavebear.com – and if I incorporate some data by reference over which I have no control, well if you have multiple parts of the domain name system, what I incorporate by reference may change without my having any ability to control it.  Well, that’s my fault for incorporating something by reference.  If I know that, I can protect myself.  And I’ve been running different route systems.  I’ve been running a non-ICANN route system.  I’ve been living in dot web and all the other tLD’s for about two years and I have not had any noticeable problem. …  A few years ago we switched from rotary dial to touch-tone telephones.  Having touch-tone telephones allowed us to have voicemail and paging systems and all kinds of new services.  Well, there’s this attitude going on inside ICANN and some standards bodies that standards are restrictive, that they form a limit beyond which we cannot pass.  And had we applied that concept to the rotary dial telephone, we would still have rotary dial telephones because there were people with rotary dial telephones who couldn’t use the new voicemail services and paging systems.  Progress leaves people behind.  HTTV is coming on now, maybe.  And you know people are going to have to buy new TV’s.  Should we say that because people are going to have to buy new TV’s we shouldn’t make progress?  Well, that’s happening with the domain name system. … Under the flag of technology, it’s being said that we must have one unified domain name system, when it’s really a policy choice.  And just like we can have multiple systems and providers for telephone directories, we can have multiple domain name services.  If you look up www.cavebear.com and you get to my machine, no matter how you did this look up, you’re happy. 

Now it turns out there are reasons to have different look-up services.  Parents want to filter out stuff from their kids.  I hate DoubleClick, for putting up all their ads, so I actually have a thing that says whenever I utter the name DoubleClick – at least in the context of fetching webpages – it turns it into a reference to my own machine where I feed it a one pixel transparent jif so wherever I would normally have a horrid DoubleClick ad, I get a blank spot on the screen, which saves me bandwidth, it saves me harassment.  Why shouldn’t I be able to control my perspective of the internet?  Why should I as the consumer of name-space be forced to consume what the provider gives me?  It turns out that there are arguments made about the cohesiveness of the domain name system that everybody should have the same experience.  If I give you a URL on the web, you should have the same experience.  Well, with things like DoubleClick and this consumer profiling going on, … if I give you a URL and I give you a URL, and you’ve both been doing things differently on the web, you would have different databases about who you are, and these commercial entities will feed you different content.  So what you’re getting is really not at all the same thing, and it’s going to just get worse. So saying that we have to preserve URLs and domain names is fallacious.  Besides, if I offer some content to you, nobody says I can’t retract it.  URLs, domain names, they’re transient anyway. 

The IETF and the IAB have been saying that, no we have to have stability.  Why add a degree of stability to something that’s not stable today?  I mean you get “404: URL not found” or “e-mail address changed” all the time.  I’m sure you’ve experienced it a lot.  So the domain name system is really just a service that one can acquire that can be subject to competitive forms of regulation. …  [D]ifferent DNS providers will be forced to provide to their customers a service that meets their expectations.  So if you look up name with their service and you get the wrong address you’re going to walk over with your feet and go with another provider.  And providers will start adding value-added services, like filtering out things you don’t want to see.  They may provide services not to you, but to providers, such as geographic proximity because right now the DNS is scattered around the entire world and your queries may be going across the Atlantic Ocean.  There’s no reason to do that.  You can actually force some geographic proximity to where you are.  There’s a whole new world of web caching and video caching going on whereby we actually put routers on ingress and egress points of your ISPs and we watch traffic coming in and out and we see you asking for something that’s way out over there.  We actually don’t let you get it.  We take apart the packets and say we have a copy closer, and we send you [that one]. 

So there’s all sorts of value-added services that can come out of playing with the naming service.  And I submit that these forces are adequate for regulation in the domain name space, with the proviso that you can’t infringe trademarks.  If somebody’s got a trademark and you are infringing – and I mean infringing in the traditional way: you are using that name in commerce in a way that infringes, you know, confuses the consumers, breaks down the differentiation or identification– then fine, let the standard legal mechanisms apply.  If a domain name is defamatory, you know if you say “John Doe is really a horrible person and cheats at business,” fine, bring a defamation action against him.  We don’t need a UDRP. We don’t need a cybersquatting statute.  And, by the way, I do not believe in cybersquatting.  I believe in speculation in domain names.  Speculation in art, speculation in land, speculation in cars, speculation in stocks.  It’s a time-honored tradition.  Morgan Stanley, a company whose whole business is speculation and capitalism, some kid beat them out to the msdonline.com.  Did they do old Adam Smith method, did they pay the guy for the thing?  No, they rattled the legal chains so the guy gave up.  There’s a place for speculation.  It’s got a good place in capitalistic society.  It brings us office buildings, it brings us investments, a lot of which fail.  Speculating in domain names …  If I have GeneralMotors.com and I’m speculating, and General Motors doesn’t buy it from me, who else is?  I have one customer, and if that customer says no, then I have to eat my investment.  And what’s even worse is that I have to keep paying my investment to keep it so that they won’t get it. 

So a lot of these arguments about regulation of the domain name space, when you look at it with a very libertarian point of view, in terms of economics, you don’t need it.  But admitted, there are a few technical design problems with DNS.  I would suggest that since any 13 year old kid a Linux machine can set up a DNS system, that if there is indeed a real threat to the stability of DNS then it should be repaired because it’s just like a security flaw. 

ICANN: Where ICANN Went Wrong

Q:  Going back to your earlier statement, you agree with ICANN in theory.

A:  Yes.

Q:  But you don’t agree with how it’s played out so far.

A:  Right. 

Q:  What do you agree with in theory?

A:  That we need some degree of regulation over the IP address space.  The other two parts of ICANN, the protocol parameters and disputes, we don’t need them.  It’s never occurred in 30 years.  There was a king of Spain some time ago who – and I wish I could find this reference –said you should not make laws about things that rarely happen.  Well, if we’ve never had a dispute over protocol parameters in over 30 years of the internet, then why do we need an expensive, heavy body like ICANN to do it?  With its own supporting organizations, as well, for just that area.  And DNS, as I have just been explaining, we don’t need ICANN over it.  So let’s just have ICANN over IP address assignments. 

Q:  So what went wrong?

A:  What went wrong.  Oh boy.  Good question.  Congress has asked and never got a decent answer.  The history of ICANN is as opposite to the Immaculate Conception as one can get.  I talked to Jon Postel a bit before he died and I don’t believe he really knew what [text omitted] was [being done] behind his back.  I believe it was being done behind his back.  I don’t know who the parties are who were involved, but I do know that Jones Day has received a windfall in terms of revenue from this.  They were involved in the creation of this organization.  They are receiving outrageous legal fees for services rendered.  [text omitted] They’ve even blown attorney-client privilege by mixing their legal representation with the fact that they’re doing policy in ICANN.  … They can’t assert attorney-client privilege.  They actually are now presumed to have broken it and would have to now re-create it. [text omitted]

Q:  What is it that Joe Sims, et al. were doing prior to Postel’s death?

A:  Postel is not an attorney.  Postel was a techie.  And here’s this whole complicated legal structure where words of art, corporate structures, governance structures …  It may not have been behind the back in the sense of he didn’t know, but Jon was not a fully informed client.  Now Michael Krieger would have better information about this than I would. …

Going from the very questionable formation of ICANN, which there were a lot of other proposals coming along.  ICANN was developed, the IANA / Joe Sims thread was done in utter secrecy.  There was a mailing list that you could send things into, but you never got any response. …  And if you ever got to talk to Joe Sims – he’s the only one who would talk in his glowering way – he would say, oh it’s already been settled, it’s too late, we can’t change it.  There’s an Athol Fugard play, “My Children!  My Africa!” and one of the lines is that the saddest words in the English language is “it’s too late.”  And that’s his [Joe Sims’] line: it’s too late to change, we’ve already decided it.  Even though you’ve [Sims, IANA, ICANN] have never heard this issue before. 

They had no notice, they had comment, but there was no consideration of the comments that anyone could detect.  So that whole thread was going on completely with its own momentum, nobody knew why.  There was this IFWP effort, which was people talking to each other.  And of course when you have people talking to each other you discover just how hard it really is to form a government.  And then came NPIA, an agency which so far has not clearly demonstrated any statutory authority whatsoever – I mean they’re hanging their entire structure off one sentence or something off one executive order that is questionable relevance and no clear statutory references.  You had to sort of be there at the time.  It was like, on yeah, the IANA proposal is in, there’s the other ones, but it’s going to be the IANA proposal.  And yes it was, it was the IANA proposal. There was no open review of the quality of the proposals.  It was just like yeah, they were the chosen child.

And then … people like Ira Magaziner, Becky Burr – mostly Ira – said yeah, we’ll make sure that they’re open.  And then Ira went and retired.  And, of course, what have we got?  The first words in ICANN’s bylaws are open, transparent, and whatever, accountable.  It says there, shall operate in the manner most open or something like that.  Have they ever had an open board meeting where you actually see the input from people coming in, see them discussing it, see them arguing about the points of view, see them counting a vote?  No, what you actually see is bad theatre.  And I do theatre, [I’m a theatre techie], so I spend a lot of time looking at actors on stage.  These are bad actors.  It’s performance art.   We’re seeing the result of what in California under the Brown act would be completely banned.  They’re deciding amongst themselves, and I don’t even believe they’re even deciding.  I think they’re mostly sitting there following Mike Roberts and Sims and whoever is saying look this way, look that way.  In fact, I don’t think that most of them, the board members, have a very good feel of their obligations as board members, … to whom they owe duties and what that duty involves. 

But getting back to the whole history of ICANN, it’s been punctuated at a very rapid rate with these indications that it is not a public body, that it rejects pubic input, that it is strongly biased in favor of commercial interests.  I mean if you look at the domain name supporting organization, they – out of the blue – said that there shall be seven constituencies.  Now if you look at them, they are like every …  Like Eskimos have a hundred words for snow, well we have a hundred words for domain name registries.  We have the global domain name registry, we have the CCtLD registry, … we have all these different variations.  Where are people on the domain names allotments?  Nowhere.  Where are schools?  They get to be a piece of a non-commercial organization.  Where are churches?  Where’s the little guy?  Nowhere.  We’ve been trying to get an individual domain name owners organization together for a long time.  Of course, we’ve had our problems, but ICANN is not even willing to, they didn’t even recognized our petition; they actually took another petition ahead of ours for a different constituency. 

They are unwilling to admit that individuals have a role.  If you look at what they’ve said all along, they’ve said ICANN can have democracy as long as individuals are not involved. We cannot trust the individual.  You know, if you look at the at-large, at-large has got this whole set of hoops that an individual has to jump through. …  You have to prove that who you are is who you are.  Do they have to do this in the commercial area?  No. If somebody walks in and says, “I’m the representative of IBM,” they take him at his word.  Do we verify that IBM is in fact a legitimate corporation?  Do we validate that IBM though its proper corporate mechanisms has given this guy the right to represent them?  Have we had this guy prove that  he is in fact … who he purports to be?  No.  We’ve given commercial groups enormous benefits of the doubt whereas individuals have to bear multiple crosses.  Hence, this is why I’m saying that ICANN has just, structurally, gone completely the wrong way. 

ICANN: Can ICANN be Fixed?

Q:  How can you … [improve the] structure?

A:  First you have to have people who don’t just say, “We’re going to be open and transparent.” … Structurally, all the pieces are there.  They have to have open meetings.  Have they had one?   No.  They have to have accountability, and I mean by accountability that we have to be able to light fires under peoples’ feet.  What has ICANN done?  It’s removed the right to derivative action.  Who can then bring a derivative action against ICANN?  The people who are members of the councils, who get to select, not elect, but select the board members.  California law allows people who elect board members to be statutory members who can bring derivative actions and a whole plethora of other rights.  Those have been blocked.  Nobody that’s affected by ICANN’s regulations has any of these rights.  The only way anybody has any authority over ICANN is to go to the attorney general of California and say, “Please, attorney general, look at this organization and do something about it.”  And those of who live in California definitely have a leg up over those who don’t.  But getting the attorney general’s attention, especially on matters as arcane as this, is quite hard.  And this is seriously arcane.

GOVERNANCE: Structuring Global Internet Governance

Q:  Do you think that any kind of internet governance structure might be better envisioned on the model of the world treaty organization, rather than as a creature of U.S. law?

A:  Probably, in the long term I think it’s going to have to be that way for things like IP addresses.  I mean, government has a place. … Government is the place where regulation, as opposed to economic regulation, where explicit regulation should come from.  It is where we have spent so much time working, trying to come out with the tensions that go in opposing directions, to give accountability and review in all the pieces.  I’m not afraid of having an international organization do it.  It couldn’t do a worse job. 

ICANN: Limits on ICANN’s Authority

Q:  What do you think of ICANN’s [argument] that their only authority is reflected by people’s willingness to subscribe to their policies and they have no inherent power to force people to do so?

A:  But they do have it.  As long as people accept the fact that there is one legacy root of the DNS – and as a practical matter that is the case, I mean everybody points to the same root and the IFWP had a chapter saying there shall be one root so as a practical matter there is only one root – they are the ones who can tell NTI what top level domains can go in the root and which top level domains can’t.  And they also regulate the business models of the registries. So there’s no options.  This is mandatory control.  The premise of your question to me doesn’t exist. 

Q:  You don’t think that there are technical means of getting around it?

A:  Well, for DNS you can start another root.  But it’s like, am I willing to go out and write another operating system for PC’s?  Am I willing to gamble everything on there being another Linux.  Maybe, maybe not. 

CONSENSUS: Defining Consensus

Q:  Tell us how you define consensus. 

A:  Oh boy.  I don’t like consensus … because you can’t define it.  When you are getting to hard decisions, I think people have to stand up and be counted because the consensus-taker – the person who is counting or measuring consensus – can become a tyrant.  And we have seen this in the DNSO.  We had the nominations … The DNSO has this structured Names Council which comes from the seven constituencies, and there’s this general assembly, a powerless group which I resigned from because of censorship.  There structure is that board seats have to be nominated by the general assembly but the actual selection occurs by the Names Council.  The Names Council members can be in the general assembly, so they can put their own nominations in.  So the net result is that the general assembly has no voice whatsoever and the Names Council can do anything.  When this happened, there were some various names put in and it turned into a … because there is no re-election, the seconding process became a stand-in for an election.  And Ni Quaynor and I both got far more seconds than anybody else, by a huge gap, considering that there were only a couple hundred people involved.  Neither Ni nor I got any votes whatsoever.  He may have gotten one.  When I got a vote, apparently it was considered a technical flaw and they had to stop the counting for three days. 

Q:  Why was that?

A:  I don’t know.  Somebody said, “Oh I didn’t really vote for Karl.”  Mike Roberts had better know that when I come in there, I am going to exercise every power given to a director under California law to review every single document that ICANN has and every process.  California law gives directors very strong authority to direct a corporation.  In fact they’re obligated to direct the corporation, and I suspect that we will find things that could very well trigger things like the IRS intermediate sanctions for 501(c)’s.  That’s a big hammer against a corporation and its board members. 

CONSENSUS: Alternatives to Consensus

Q:  So what’s a good alternative?

A:  Votes.  Clear votes.  Clearly articulated issues with clearly counted votes by a clearly enumerated electorate. 

Q:  And who should be included in the electorate?

A:  The electorate should be ultimately anyone who is affected by the decision.  We have the problems of cost and we have the problems of identifying people.  But to me that’s better than this fuzzy consensus thing. 

Q:  So if we have votes, would you want to use a majority standard?

A:  It depends.  Majority voting can be a bad thing.  Staggering board seats is a classic way for keeping the minority from getting a pro rata share of the board seats on a corporation.  Cumulative voting, multiple seats at the same time, cumulative or STV voting are techniques that tend to cause representation of the board or whatever that roughly approximates the interest distribution.  And you can see here that we’re not in a situation where we have two groups who are just trying to adjust little details.  We’ve got diametrically opposed points of views on a lot of these things.  So if you look at the at-large, again, you can see that the elections have been staggered, that there’s majority voting – I think there’s majority voting – but there’s only usually one seat available so it doesn’t matter whether there’s cumulative or STV voting since there’s only one seat …  That means majority rules everything.  And you have all these elections and the majority wins and the majority wins and you end up with a board that’s entirely majority selected or the decisions are entirely majority selected and the minority, even if it’s 49.99%, never gets anything ever. 

GOVERNANCE: Policy Questions vs. Technical Questions

Q:  I want to loop back to something you explained earlier about the distinction between policy questions and technical questions and how there’s not necessarily a clear line.  In a kind of system where you’re having a determinate vote with an at-large membership however it’s defined, can you talk about the kind of technical questions that are appropriate for at-large voting and the kinds of things that should be determined by the standing body?  What’s the distinction?  What kind of questions come to the fore?

A:  I think ultimately the entire group should have the authority to look at any matter.  And that doesn’t mean that they can’t at some point say that, somebody should put together a proposal that says, there are certain classes of matter which we believe … we can put certain limitations on the exercise of discretion about those, that we are willing to hand off to an executive staff or circumscribe the board in some way.  Like for example in the Boston Working Group, when we put together our submission to NTIA parallel to the IANA one, we actually had a thing that said that ICANN has these fundamental assets – you’ll notice I like the word fundamental – and they consist of the – I can’t remember the exact language – basically the rights that they’re being granted by the government to control these assets.  And these are rights to the root zone, IP addresses.  We said ICANN can do the following things with them, but it cannot alienate them, it cannot regulate, it cannot license their use for more than a period of so many years.  We drew a line around it.  That’s the sort of thing that an elected body can do.  It can pass a motion that says, we will give you the board, or you the executive board, the ability to make these decisions as circumscribed. And that’s classical management by delegation.  Draw the borders around what somebody can do, but within it they have discretion. 

Q:  Can you give some examples of, if you were to draw those borders, what would fall on either side of the line?

A:  Well, so far since the general public has not had a chance to deal with any of these matters, I would take them all out of ICANN’s hands.  I would want everything that is decided to be reviewed and decided on de novo. 

GOVERNANCE: How Would You Govern the Internet?

Q:  Assuming you get past that point where you’ve put together an institution that’s open, then can you give us a sense of the issues that you would delegate and that you would keep.

A:  So if I were the emperor of the universe? … Well, first of all I would take DNS out of ICANN because as I’ve said I think there are other forces that can handle that.  But with respect to IP address assignment, I would delegate to the three existing regional registries the ability to continue registering addresses within those zones.  However, I would also appoint another group that would look to make sure that these three regional groups – ARIN, RIPE, and APNIC, are you familiar with that?  They’re the Asian, European, and U.S. or American address assignment registries.  They roughly correspond to the big land masses on the planet.  They also correspond to the clumps of interconnectivity on the internet.  There is basically lots of interconnectivity within them, and sort of thin connectivity between them.  Which is real nice for doing your address allocation system.  Now, if suddenly Africa should become a big, solid, thick clump of connectivity with sort of a few links to Europe, it would make sense at that point to split the European registry to an African and European registry.  So currently, I would say RIPE, you’re doing a great job, but I reserve the right to split you.  Or aggregate you.

ICANN: The Worst ICANN Can Do (Part I)

Q:  What’s the worse thing ICANN can do?  In terms of impact or specific action?

A:  It’s setting a model for internet governance, for other areas that will arise.  And it’s building the cookie-cutter of the future.  And it’s quite an undemocratic cookie-cutter.  Yet people are quite willing to copy it.  That’s what scares me the most.  DNS, when you get down to it, is kind of boring.  I’m probably going to be able to keep auerbach.com although people have threatened to take it away from me.  But, it grows.  If the area it covers grows, and ICANN itself is threatening to grow in other areas.  Hans Kraaijenbrink – I cannot pronounce his last name – has completely ignored the address supporting organization and has created this ad hoc group on addresses which appears to have no purpose other than to deal with address issues in a much broader context than IP addresses.  It’s looking to IP addresses and telephone numbers, and that kind of expansion in a similarly undemocratic sense – in fact in a worse undemocratic because at least ICANN’s current structure says that the board shall delegate to the supporting organizations, even though it doesn’t do it – here’s the case where it’s creating a structure where it doesn’t even delegate to the supporting organizations.  So what I’m afraid of is it [ICANN] growing.

THE FUTURE: ICANN In 10 Years

Q:  Look into the crystal ball, ten years from now, what is ICANN’s scope of authority? What functions does it serve?

A:  Well, I think DNS is going to return to what it should be in the first place, which is a system that hands out handles that map into IP addresses and IP telephone numbers and mail exchangers, you know because DNS covers a whole number of things, not just web access.  And these names will rarely be seen by humans.  Most of the time when you, as a human, deal with the internet you will deal with some intermediary that knows how you want to deal with the network.  When I say “Chris” I mean my wife Chris.  If you say “Chris,” you mean someone else.  But your access tools to the net will know your context.  And when you’re looking for something, it will know your preferences and things will be tuned to who you are, and you will not be typing domain names except in very rare circumstances.  About as often as you type in IP addresses now.  And to that end, I’m hoping that DNS will submerge to become just the internal gear, just like a piston in your car: you know its there and you hope it works, but you don’t see it and you don’t really care what color it is or if it’s domed or if it’s dished or whatever.  It just does its job.  And therefore all this heat and smoke and stuff will become irrelevant, at least to the DNS.  And IP addresses are so arcane anyway that you won’t see them in a few years.  That’s going to become so specialized that it’s going to have its own, like the people surrounding telephone exchange, there’s a whole legal profession built around how telephone companies exchange traffic with each other.  That will become, I think an arcane little niche and ICANN will fall into that and be that.   

THE INTERNET: Commercial Interests

Q:  I want to go back to something you said earlier … In your opinion how has the growing financial stake in all of these questions affected the debate about internet governance?

A:  The DNS space is quite overt.  The trademark lobbyists have been quite strongly interested in protecting their rights.  And they’ve shown that they are very efficient in organizing and lobbying and what have you.  The economic stakes on the IP address side, most people don’t realize it.  Some companies realize it.  Cisco realizes it.  When Hans came out with that ad hoc group, the IETF, the three RIRS, and Cisco systems all sent in highly negative comments about this.  Comments that were ignored.  So the economic interests of those who in DNS, have the DNS trademark interests have clearly floated to the top and are very visible.  The economic interests of those who would use domain names for other purposes are quite diffuse.  I mean, what is my real economic interest in having my own last name as a domain name?  It’s hard to put a number value on it. … So cumulatively – I may have a smaller personal stake – but cumulatively it adds up to a possibly much larger interest – I don’t know how we’ll measure it – than the trademark interests. 

So the focused economic interests have floated to the top and have made themselves clear in the domain name space.  The registries are clearly trying to exercise power because there’s money to be made.  Everybody’s looking at NSI and saying look at all that money, I want that and do that too.  But in the IP address space, it’s just such an arcane issue.  From the point of view of Cisco, we build routers, that’s our life.  It’s how do you handle IP addresses, that’s critical to our business that they don’t break – the fundamental attributes of how we have packets that move around the net.  So yeah, economic interests are making themselves known, but there are not a lot of people who know they have an interest to articulate it.  That may be a better way of answering your question.  Not everybody knows that they have an interest. 

ICANN: The Worst ICANN Can Do (Part II)

Q:  Right.  That seems to be the big danger of what ICANN is doing.  People don’t realize that ICANN is making decisions that will affect them for a long time. 

A:  And a lot of times since you didn’t know any better – you don’t know that the grass is greener on the other side since that’s the way you’ve always lived.  If you always lived in a coal mining town in South Wales, you may not realize that you don’t have to live with black lungs.

THE INTERNET: What is the Internet’s Greatest Promise?

Q:  I guess one other big picture question – it’s the one we usually start off with: what’s the internet’s greatest promise.

A:  If I want to say it in new age, then empowering individuals.  I don’t know where it’s going to go.  I just think it’s a great ride.  It gives me the ability to find people who I’ve never dealt with before.  It let me find my missing half-brother.  To me, that’s like incredible.  I mean he disappeared when he was only three or four years old and I never saw him, even when he could speak.  And I got an e-mail one day from a person looking for my grandmother.  He mentioned that it was his grandmother; we’re half-brothers.  It’s that kind of thing that the internet is enabling.  It’s incredibly valuable in an unmeasurable way.  It really helps.  Computers are really good at doing things by rote.  People are really good at doing arbitrary things, leaping to conclusions, creating new things.  The internet is really a great tool for getting two people, who have each the half of an idea, together so that they can create something new. 

Now that’s its real promise.  How it does that, we don’t know, but we shouldn’t stop the innovation.  And by regulating the internet, when you put a regulation around something, you actually stop its innovation.  And by stopping innovation of domain name space, we are stopping innovation of how you name things.  How you name things is quite a strong tool in how two people will find each other.  I mean, my half-brother found me because I had auerbach.com.  If I hadn’t had that, he would never have found me and I would never have met him, possibly.  So that’s the value of the internet.  What was the power of the book?  Guttenberg didn’t know. 

GOVERNANCE: Approach

Q:  Do you know them [Mike Roberts, Joe Sims] on a personal level?

A:  No.  I know many other people, but I don’t know Mike Roberts. … Actually, I hear people say really good things about him, but not in this context.  He had to be a real personable person to sway Jon Postel.  Jon was not a dummy.  So when I said things were done beyond his back and Jon wasn’t necessarily particularly knowledgeable, Jon is smart at learning, but he is definitely a techie.  And one problem we techies have is that we tend not to be educated well in disputes and soft issues.  We’re not very good at it.  In fact, I’m afraid of them.  I can argue with someone about a technical issue until we’re both red in the face, and nobody feels the worse for it.  A knock-down, drag out technical fight, because there’s a good, clear answer.  But in this soft stuff, most of the technical people, first they don’t like it because it’s hard to come up with clear answers.  But everybody takes it too personal.  There’s not enough of a sense of humor in all this. … There needs to be more people laughing …  This really isn’t all that important.  Nobody is going to die over this.  People are starving.  This is not on that type of scale.  Let’s all relax, take a break, come back, shake hands, and start all over with things.

PERSONAL BACKGROUND IN INTERNET ISSUES

Q:  Can you say a little bit about your background and how you got involved in the debate, that kind of thing?

A:  Well … my family has a history of radicalism.  Being arrested for various protests is a mark of honor in my family.  Unfortunately, I do not have this mark of honor.  I have never been arrested for a protest.  I’ve been beaten up by the police.  So, the long history of being troublemakers, and asking questions, questioning authorities, has been ingrained from day one. …  Background?  Normal, Van Nuys, California.  I went to Van Nuys High just like Jon Postel and Vint Cerf. … They were there a few years before me.  Although I met Vint Cerf in 1974, he was a consultant to my group.  And had normal educational experiences.  Went to UCLA and studied physics and what have you.  And I discovered I have no mathematical talent whatsoever, which is a bad thing for a physics major.  At this time there was no such thing as computer science, although computers fascinated me.  So I took the easiest major I could find Geology, which I took all engineering classes.  So I somehow developed this split in my brain between technology and soft stuff.  And I’ve had that ever since, which is why I spent half my life doing things like theatre, and then I go do technology.  I have not an ear for foreign languages.  I’m the only person in my family who can’t speak a dozen languages, so I couldn’t get out of UCLA, so I transferred to Berkeley, which didn’t have a foreign language requirement.  Graduated from there.  Then did the obligatory getting tear gassed a couple of times in People’s Park. …

Moved back down to L.A. and got a job with an aerospace company with security clearances and all that sort of nonsense.  Then one night, got pulled over by the L.A.P.D. or the L.A. Sheriff’s and my car got searched, and I got searched and my girlfriend got searched.  Everything got searched.  And I said, there’s something wrong here.  I want to know more.  So someone I knew threw a law school catalog at me.  I applied to exactly one law school, took the LSAT, got accepted.  I was this long haired, furry techie person.  I did reasonably well in school, graduated like third or fourth in my class – this is Loyola Law School.  And never practiced.  I went back into technology and looked for a worked for a long time on Unix types of things.  Started my own start-up companies.  First one was extremely successful.  Unfortunately, I gave it away in a divorce settlement.  Second one a total disaster, so I now know the feeling of a business is going south in a hurry.  The third one I started, it got acquired by Cisco, so I now know the feeling of one that’s very successful. …

I’ve always been interested in law and social issues.  Always reading, watching, but I’ve never really done much with it.  But for the last 30 years I’ve been building the internet and when IETF started drifting into policy issues, of course I got involved because this was naturally a magnet for me.  So I got involved in the IETF policy group, trying to figure out how the IETF should run itself.  Got involved in the so-called IAHC; I don’t remember the initials, something ad hoc committee to do the first things with domain name stuff, although I wasn’t on the committee at the time it was doing this stuff.  I’ve just been following this whole internet governance thing since several years.  You know, when ICANN got started, not started but when the discussions got going, there was this IFWP, hopped on this flight and went to Reston, had a great time, met a lot of great people, got a lot of great ideas.  And as the IFWP process went along, it turned out that I had bought tickets to come here [Boston] for the IFWP meeting and the meeting got cancelled.  There were a dozen of us who had tickets, so we said let’s go to Boston anyway.  And that’s how the Boston Working Group got formed.  Diane [Cabell] was our scribe.  And we wrote up our own counter-proposals, which are somewhat worth reading because it was a lot of good stuff.  We took the IANA proposal and we said … it’s the locomotive, it’s going, it’s not going to be derailed, so let’s see what adjustments we can make.  We proposed some good adjustments, they were virtually all ignored. …

When ICANN got formed and I saw the initial board I was actually very pleased. So I said let’s give it tried.  Then disappointment, disappointment, disappointment, disappointment, to the point where I’m now just downright angry that it continues to exist.  I think it’s done a great job of creating the 1954 government of Bulgaria.  It’s a true Soviet, responsive to itself and no one else.  That’s how I got where I am. 

You know I have an entirely different life.  I’m in the CTO's office at Cisco, and I’m working on where the internet is supposed to be from an infrastructure point of view.  I spent five years building systems to move broadcast quality and better video around the net.  So my perspective of the net is very different from most people.  I don’t see the net as web, the world wide web.  I see the net as where we move packets around for mult-media or all kinds of other new things and a lot higher data rates than we’re seeing.  So I see technology which some of it will never see the light of day, and which most of us will be seeing in five, ten years.  So my perspective on ICANN is quit regulating that, and let’s deal with the future and the present, but my present is most people’s future.  At least, I hope it’s in their future.

 


Search:
BCIS • POUND HALL 511 • 1563 MASSACHUSETTS AVE • CAMBRIDGE, MA 02138 • T (617) 495-7547 • F (617) 495-7641