[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [ga] Older registrations



Craig,

Time to reminisce again Harald... ;-)

>Simon,
>
>I want to congratulate you for providing a very interesting and useful
>recounting of these events. Folks here may be interested to know that also I
>have some stuff written about this at
>http://www.flywheel.com/ircw/dnsdraft.html and

Um... thanks, but I've definitely been misquoted. I think I have to put a 
hit out on you now... ;-p

RE: (This quote awaits verification)

I questioned whether it would be right for me to be on the first IAHC 
because I was also a TLD applicant and it would probably constitute a 
conflict of interest. Jon thought it would be OK if I just sat out my 
application (it was supposed to be a less formal process than today). But 
before I got a firm answer back ISOC had taken the process away from Jon. I 
only described IAHC in negative terms after their first draft when they had 
proved to me beyond a shadow of doubt they were not listening to, or 
benefiting the Internet community, or even attempting to build on the prior 
IANA work. What I actually said will probably get your web site RBL'd. ;-)

If you could amend that paragraph please.

>http://www.flywheel.com/ircw/overview.html .
>
>Of course, you were there through the whole thing, and I was not until 
>spring of
>1997 (and it took me another few months after that to figure out why the
>participants seemed to hate each others' guts so thoroughly). I am primarily
>interested in assessing this in terms of the wider context of modern 
>society and
>social thought: namely, the big picture story about how computer engineers
>engage in social architecture. But it's also important to have a clear view of
>how the smaller pieces of the story fit together.

Just an FYI, I was the recipient of the CommerceNet '96 VIP Award for Best 
Online Community for a site I designed in 1995. The technological 
architecture definitely shapes and enables social architecture if done 
correctly. This comes down largely to correctly programming the human 
interface, but other areas are equally instrumental to the human psyche 
such as "identity". This, of course, is where names and addresses become 
important.

Having said that, the best online resources showing computer engineers 
engaging in social architecture are http://www.dilbert.com/ and 
http://www.userfriendly.org/. ;-)

>That said, I want to ask some questions and add one or two comments.
>
>1) You said you opened up the original shared-tld mailing list. I'm afraid I
>don't seem to have anything on that one. Can you point me to archives?

The mailing list was never archived. I have a copy of all the traffic in 
Eudora mailbox format on a CD, but no-one has ever asked to see it.

>2) You mentioned a plan to create a small series of ad hoc committees. 
>This is a
>very important point, mentioned in the January 96 draft at
>ftp://rg.net/pub/dnsind/relevant/draft-ymbk-itld-admin-00.txt . But the 
>authors
>were floating ideas. The same draft contemplates the alternative, creating a
>more formal permanent body. What's your sense of who favored a permanent 
>body at
>the time, and who preferred the ad hoc approach?

Who has motive and who benefits? I think everyone involved in the early 
days favored independence from governmental oversight. The idea of a 
permanent body was to establish a stable mechanism (something that would 
have the authority to delegate new iTLDs). Unfortunately, there are people 
who have used the idea of a permanent body as a career leveraging move (a 
marketing decision). Thus the establishment of a blue ribbon panel instead 
of a panel of unwashed geeks with rough consensus and running code. Looking 
back now writing this I think this is how the original situation got 
thoroughly "Dilberted".

>3) One key point of controversy that I think needs emphasis was Postel's
>motivation in finding alternative funding sources for the IANA and the
>consequence of this. Some of the Spring 96 drafts contemplated a hefty fee for
>blessing a new TLD registry, and this provoked heated opposition.  Isn't this
>what prompted Denninger, Kashpureff and Fenello to try to go out on their own?

Each of their motivations were different. Karl didn't agree to any kind of 
taxation. Eugene was Jon's grand experiment which went wildly out of 
control (in another Jon-prompted situation I ended up enquiring on buying 
NSI from SAIC and discovered SAIC's plans to IPO it). I'm not sure about Jay.

The "hefty fee" was a 2% royalty per 2LD delegated, paid to IANA. I think 
most agreed with this in principle. The problem was defining the audit 
trail. There was a lot of uncertainty among the competing registries about 
how the audits could be trusted, and the misunderstandings led to this 
royalty being called a tax. Remember that we were designing a system to 
create competition against NSI and creating a commercial advantage is part 
of that, whether you like it, or not. This is probably where the claims of 
"greed" come from. CORE, on the other hand, created an environment of pure 
greed (a closed, exclusionary community) and sanctified it by doing it in 
the name of a public trust. I've stood at CORE members trade show booths 
and have been plain lied to. Fortunately they've now been roped in by USG 
and are now having to compete legitimately with the rest of the Internet 
community.

>4) Ambler, on the other hand, recognized the potential windfall of obtainin an
>official blessing that he could own .web, and he was more than willing to 
>pay to
>get it. Thus, as I see it, he tried to force a contract on Postel by way of
>Manning. This is where personality makes a difference. Postel was by this time
>apparently rather fed up and didn't want to be bothered; Ambler is one of the
>most aggressive people you'll ever meet; and Manning is the archetypal 
>nice guy
>who is uncomfortable saying no to people. He said no to Ambler's entreaties at
>least once that fateful day, but Ambler kept up the pressure, and (with
>everybody buzzing on the MSG from too much Chinese food?) Manning finally
>succumbed, accepting a sealed envelope that Ambler calls an application 
>fee. At
>least Manning kept enough good sense to sign nothing. Postel intervened 
>the next
>day, sent back the envelope to Ambler, and announced that no commercial
>applications were being accepted. (Unfortunately, I just discovered that the
>link at umich.edu that I used to cite Postel's email to this point is no 
>longer
>good).

I was there too. That's one opinion you could make if you weren't there. ;-)

First, Jon Postel was never employed by IANA full time. He was employed by 
USC and his duties included mostly IANA-related things. He was busy on 
other projects that day. Bill Manning arranged the meeting as a consensus 
gathering exercise for Jon's draft. I never figured that out until the end 
of the meeting and we agreed upon the notes/minutes for the meeting. Both 
Chris Ambler and myself attended as TLD applicants. During the meeting Bill 
got out the file of applications and confirmed our TLD applications were on 
file. The check is a really big red herring. Because the Postel draft was 
going to require a fee to accompany the applications before they could be 
processed, Bill agreed to add a sealed envelope to the applications with 
the understanding that the contents were to remain sealed until the 
application was processed. The purpose of this was so IANA would not 
receive money ahead of the application, yet the application could be 
immediately processed when the proper time came. In practice it would work 
out the same as if we were to go home and mail in the check based upon the 
Postel draft requirements. Either one would be satisfactory. Remember, the 
underlying motive of all this was to satisfy the Postel draft requirements, 
have the iTLD application successfully processed, and get the chosen iTLD 
in the root. I didn't have my check book with me otherwise I would have 
done exactly the same thing.

Did Bill Manning or Jon Postel know what was in the envelope? I know who 
knew, but it really doesn't matter one way or the other. They weren't doing 
anything wrong. What caused the stink was that someone leaked the rumor on 
newdom that IANA was accepting payments for iTLDs. This wasn't true. The 
facts are that IANA did not, at any time, receive money for the delegation 
of iTLDs. But in order to avoid the appearance of doing so - now the whole 
world knew what was in the sealed envelope by way of newdom - they were 
forced, by convention, to publicly return the envelope.

I can't comment on Chris' claims as I was not part of his private 
discussions with IANA. What I can comment on is IANA's blessing at that 
meeting for the proposed registries to fulfill the requirements of the 
Postel draft. This included turning on the registries to meet operational 
requirements, and ensuring the visibility of the iTLD's DNS records to the 
public. This also included accepting paid registrations in the applicant's 
chosen iTLD(s) - I was against collecting a fee until the iTLD was in the 
root but Bill said it was part of the overall operation so it was OK. Later 
Bill emailed out a legal disclaimer for the test registries to use that 
explained that domain name registrations would not be in the root until the 
iTLD application had been approved, and that that may, or may not, happen.

>5) I don't know how much CORE pressured Postel to reorient part of the root. I
>do know that Vixie and others had been pressuring him for some time (years) to
>take control of the dot, because they didn't trust NSI. If CORE (or CORE
>supporters) were pressuring Postel in early 98, which certainly seems 
>plausible,
>it is clear that who ever was doing that was simultaneously escalating that
>popular old anti-NSI rationale, offering scenarios that NSI was on the 
>verge of
>adding its own selection of private TLD operators to the root. It would be 
>great
>to know more about the evolution of those rumours, and Postel's private
>reflections about what happened.

I can't give you Jon's private reflections. I can only give you my 
observations. Jon was pressured from the time AlterNIC came online because 
no one had ever challenged the root authority before on this scale. Paul 
Vixie (of BIND, ISC, and f.root-servers.net fame) and Eugene Kashpureff 
(AlterNIC) went head-to-head at the 36th IETF (June 1996). I think it was 
there that Paul coined the phrase "DNS pirate". Think about it. The root 
server operators jealously guard their Internet wizardry and prowess in 
holding the net together. This was directly challenged by a the creation of 
a second experimental root. What if that root became more popular, and AOL 
and the ISP community started to use it instead? That pressure increased 
from all sides until the original process was taken away (announced by Don 
Heath at the IAHC BOF at the 37th IETF). Now look, today, at who is locked 
in, economically, as registrars to the ICANN root: 
http://www.icann.org/registrars/accredited-list.html Is the light bulb on yet?

Also Jon was worried about finding a legal umbrella for IANA's activities. 
Liability coverage was an issue that proved to be an obstacle to the 
original Postel draft process (the contested 2% royalty would have funded 
legal expenses). There was the increasing involvement of the ITU, which 
could have provided that umbrella, at the expense of the loss of both IANA 
and the IETF's independence. The ITU is well known for assuming control of 
communication standards and then charging for documentation, and this was a 
disturbing thought for the keeper of both the Protocol Numbers and RFC 
Editor. So far these particular concerns have proved groundless but at the 
time nobody (including Jon) had the answer. When ISOC took the process away 
from Jon, IANA gained that legal umbrella. I think Jon traded the original 
iTLD process away to ISOC for that legal protection in order to prevent a 
wholesale takeover by the ITU. He was a consensus builder which ultimately 
means he went along with whatever the prevailing community wanted, even if 
(as previously documented) he really didn't want to go in that direction to 
begin with. This might explain his actions in splitting the root. It was 
just "an experiment" remember.

NSI has always been expected to add any new TLDs to a.root-servers.net, but 
only from the master zone file held by IANA. NSI were contractually bound 
via the NSF agreement to IANA's ultimate authority for the contents of the 
root zone. The rumors of NSI adding private TLDs came from the same people 
who took the original process away from Jon Postel - who just happened to 
later decide the name space was a public trust and, later still, attempted 
to reorient the root. Opportunity and motive. I don't really know why Jon 
did what he did, but by that time IANA was a willing participant in the 
gTLD-MoU. The reorientation itself showed that the lines were clearly drawn 
between the gTLD-MoU and the US Government.

>6) At the time it seemed like he had to eat crow, but the consequence has at
>least one healthy aspect. If there can't be some adhoc process controlled 
>by the
>IETF/IAB/ISOC crowd, it won't be controlled by NSI, or by the USG acting
>cavalierly either (as it did when NSF first allowed NSI to start charging).

The US Government was not acting cavalierly by allowing NSI to charge. 
While I may not agree with how it was done, it was a legitimate 
cost-recovery exercise to remove the burden from the U.S. taxpayer from 
funding domain names in the international space. The contractor involved 
still needs to get paid for processing the applications. How this is done 
is where it gets messy.

>Postel's move forced the USG to pay even closer attention, and to tread
>carefully,

The USG took note of the lines of polarization and intervened directly. 
This resulted in the complete invalidation of the gTLD-MoU by the USG 
(according to Ira C. Magaziner in a conference call). It also accelerated 
the creation of ICANN.

>and it forces everybody else involved to seriously consider what it
>means to have a legitimate formal process for governing key elements of the
>Internet's most hierarchical feature. Even if this means endorsing some 
>kind of
>proprietary pioneer preference for gTLDs (which I don't prefer, sorry), at 
>least
>and at last the rule would be openly and widely stated.

It's a non-zero sum game, and the sum of the whole should be greater than 
the sum of individual participants. That should make everyone a winner 
(pre-IAHC, post-IAHC, and post-ICANN). That's how you build a successful 
community. Not by forming a Draconian empire and enforcing totalitarian 
control - it won't work as the Internet will route around it (again).


Best Regards,

Simon

--
The future is still out there...

--
This message was passed to you via the ga@dnso.org list.
Send mail to majordomo@dnso.org to unsubscribe
("unsubscribe ga" in the body of the message).
Archives at http://www.dnso.org/archives.html