Reconsideration Request 00-11
Received: 15 December 2000

INTERNET CORPORATION FOR ASSIGNED NAMES AND NUMBERS
RECONSIDERATION COMMITTEE

REQUEST FOR RECONSIDERATION FROM
DENIAL OF APPLICATION FOR NEW TOP-LEVEL DOMAIN

I. Introduction.

Pursuant to Article III, Section 4(a) of the Bylaws for the Internet Corporation for Assigned Names and Numbers ("ICANN"), Sarnoff Corporation ("Sarnoff" or "Petitioner") respectfully submits this Request for Reconsideration ("RFR") from the denial of its application for the delegation of a new top-level domain. This RFR is not intended to reargue the merits of an application or present the Board with new facts that were not already before it at the time it made its decision; rather, it is brought to correct two critical mistakes, one factual and one procedural, that led directly to the denial of Sarnoff's application. Because the mistakes made with respect to the Sarnoff application were unique to Sarnoff, the relief requested in this RFR should have no any impact on any other TLD application presented to the Board at the November, 2000 meetings in Marina del Rey, California.

The factual mistake addressed by this RFR was a misunderstanding of the effect of the Sarnoff and NeuLevel partnership on the applications pending before the Board. As Sarnoff will demonstrate, the respective applications submitted by it and NeuLevel had not been changed in any material way after the announcement of the two applicants' affiliation. Based upon contrary information delivered to the Board by its counsel, who was not aware of the details, the Board misunderstood the parties' arrangement in apparent reliance on the incorrect information. The procedural mistake was a deviation from the criteria established for the evaluation of TLD strings. Rather than review the applicant's proposed TLD strings against the objective evaluation criteria previously published for the benefit of the applicants, the Board evaluated the aesthetic qualities of Sarnoff's proposed TLD strings.

The Sarnoff application had passed all staff and Board reviews prior to the time that these errors were made. Sarnoff believes that its application would have been accepted had the Board been presented with the correct facts and followed its prescribed evaluation criteria. Such objectively apparent mistakes in the decision-making process are the very things that ICANN's March 4, 1999 Reconsideration Policy was designed to correct. All of the assets, personnel and resources that were presented to the Board for fulfilling the Sarnoff application stand ready still to bring a new personal domain name registry to the market. Accordingly, Sarnoff respectfully requests that the Reconsideration Committee recommend to the Board that it amend Board Resolution 00.89 to add Sarnoff's TLD application.

II. Request for Reconsideration.

A. Contact Information.

All correspondence, questions and responses to this RFR should be directed to the undersigned counsel, with a copy to:

Shailendra Suman
Sarnoff Corporation
201 Washington Road,
Princeton, NJ 08540
Telephone: 609-734-2000
E-Mail: ssuman@sarnoff.com

B. The Specific Action for Which Review and Reconsideration is Sought.

Sarnoff seeks a review and reconsideration of Board Resolution 00.89.

C. The Date of the Action in Question.

The resolutions in question were passed on 16 November 2000 at ICANN's Second Annual Meeting in Marina Del Rey, California.

D. The Manner in Which Petitioner Will Be Affected by the Action.

ICANN's decision not to include Sarnoff in the final negotiations toward execution of appropriate agreements between ICANN and prospective new TLD registry operators will have three substantial, negative effects on Petitioner:

(1) Sarnoff will be precluded from entering the new, global market for TLD registry services directed toward individual domain name registrants until an undefined future time, while one competitor, Global Name Registry, Inc. (".name"), will be provided a monopolistic grant to tap this new, global market, unchallenged by competition. At such time as Sarnoff is allowed to enter this market, if ever, it likely will find an entrenched competitor with a dominant market share.

(2) During the undefined period in which it will be kept out of the global market for TLD registry services directed toward individual domain name registrants, Sarnoff will, under all reasonable scenarios for numbers of registrations, lose significant revenue and business opportunities.

(3) Prospective registrants for personal domain names will be deprived of choice and the benefits of competition when selecting a global TLD in which to register.

E. A Temporary Stay is Requested.

In order to allow the new market for individual domain name registrations to be opened to all newly accredited individual domain name TLD registries at the same time and on an equal, neutral basis, Petitioner asks the Board to make no recommendation to the Department of Commerce on additions to the root zone file until this RFR is resolved.

F. Relief Requested.

Sarnoff respectfully requests that the Reconsideration Committee recommend to the Board that Resolution 00.89 be amended to authorize final negotiations toward execution of appropriate agreements between Sarnoff and ICANN for the operation of a new TLD.[1]

G. Grounds on Which Action Should Be Reversed.

The Board's decision to remove the Sarnoff application from further consideration as a new TLD operator was based on (1) a critical misstatement of fact that the Sarnoff and NeuLevel applications had "changed dramatically" and (2) a departure from the established evaluation procedure during which the Board measured Sarnoff's TLD string against a previously undisclosed criterion that the TLD string be "pronounceable" and/or a "mnemonic" for some longer word or phrase. Because the Sarnoff application received a strong evaluation in the Staff Report, received favorable comments throughout the evaluation process from both members of the public and the Board, and was the last application removed from consideration during the annual meeting, Sarnoff believes that had the Board correctly understood the nature of the Sarnoff-NeuLevel joint venture and remained true to the published evaluation criteria, the Sarnoff application would have been included in Board Resolution 00.89.

1. The Review of the Sarnoff-NeuLevel Applications.

In order to understand the effect of the mistakes noted above and how they led to the denial of Sarnoff's application, it is important to first review the Board's evaluation of the Sarnoff application as it was discussed throughout the November 16, 2000 meeting. In an iterative review that narrowed the field to a handful of qualified applications, the Board discussed every application before it, in the context of the categories set forth in the Staff Report of November 10, 2000 (hereafter "Staff Report"). As the quoted provisions of the transcript reveal, on two separate passes through the applications, the Sarnoff application was reviewed favorably, with no concerns or reservations expressed about its naming convention, its technical infrastructure, its protections for intellectual property, or its business plan. When the Board first reached the category of "personal" domain names, it had only favorable things to say about the Sarnoff application:

TOUTON: Are there any other entrants in this group [for personal domain names]?

CERF: First, a clarification. Since we had said that JVTeam was a strong player, we now have JVTeam and Sarnoff together. What.. does that do anything funny? Does that mean we have implicitly put JVTeam/Sarnoff group into the basket as a potential player or not?

TOUTON: Well, I think it is up to you to put them in. As I understand, the process that was discussed before, you put things into the basket and see if there are too many of the same color.

DYSON: I would propose putting them in.

CERF: I think they represent a strong proposal, especially in combination. So, I would recommend including them. And I have a couple of other suggestions as well, but we will wait and see when this one comes.

TOUTON: Comments about JVTeam Sarnoff?

FOCKLER: I would support it too. I am looking for perhaps even some competition in this area. I would just like to point out what we are talking about in that particular case is dot i i i.

TOUTON: Yes. Any other comments about JVTeam Sarnoff? OK. I sense a feeling that it should go into the basket.

See, Transcript of Board Meeting of November 16, 2000, at p.7, ln.18 – p.8, ln.21 (copy attached as Exhibit "A") (hereafter "Transcript," with page and line references noted by "<page>:<line>").

Over two hours later, the Board again reviewed the Sarnoff application, this time in its second pass through the applications:

TOUTON: Alright. I understand that on the list as a whole, there is a general comfort level. I think perhaps formally we should go through each one and figure out if there is a basic consensus behind each one keeping in mind what the whole list is... and the desires of some of us to make sure there is diversity, geographically, of type. There isn't the same operator appearing too many times and those sorts of things.

[Board reviews applications]

TOUTON: The Sarnoff JVTeam Sarnoff personal name super joint venture. Everyone...

CERF: It's OK.

TOUTON: Everyone OK on that? Nod of heads please. OK.

See, See, Transcript, at 13:24 – 14:17.

The first objection voiced about the Sarnoff application came approximately twenty minutes later in a discussion about what specific TLD strings to delegate to the given applications. In a discussion in regard to the Afilias application, Mr. Kraaijenbrink had made the point earlier in the day that each application accepted should specify a TLD string. Following on that point, Mr. Roberts returned to the Sarnoff application to object to the selected string:

ROBERTS: Going back to Hans's point about the delegation of strings, I would like to personally say that I find the triple i thing not acceptable as a string for the TLD space.

WILSON: Talk some more, Mike...

ROBERTS: Yeah... it's essentially unpronounceable. It has no mnemonic value. It is confusing. It says nothing about what that might actually mean.

See, Transcript, at 14:22-15:6.

Seeking to address Mr. Roberts' stated concern, several Board members discussed whether to delegate an alternative string, specifically mentioning the strings ".per," ".self" or ".ind." See, Transcript, at 15:8 – 16:5. The conversation concluded with a recommendation by Mr. Roberts to delete the Sarnoff application from consideration altogether, based solely on the choice of the TLD string:

TOUTON: Any other comments on the string for the Sarnoff JVTeam?

CERF: Well, I'm not sure how we ended up on that one. If P E R is clearly a potential problem because of the nation of Peru, my reaction to your comments is not to feel too differently than you do, but on the other hand if that was the proposal and that was their choice, then perhaps we can just let them go with it and see what happens.

FOCKLER: Agreed.

MALE VOICE: Yeah, I could live with dot triple i.

ROBERTS: Well, I respect that. I mean my reaction would be to delete it from the list.

TOUTON: I am sorry. Your reaction would be to delete the entry from the list altogether?

See, Transcript, at 16:7-25.

Leaving Mr. Touton's question unanswered, Mr. Kraaijenbrink moved to end the discussions and approve the eight applications that had passed the Board's review, including the Sarnoff application.[4] Although there appeared to be consensus on the eight applications that had been approved thus far, including Sarnoff's, Mr. Touton asked the Board to review them one final time "to be sure we have consensus and it's not misread." See, Transcript, at 20:6-7.

The Board moved through the applications again, one by one, pausing for discussion only twice. The first substantive discussion of an application in this round came in the context of the Afilias application for ".web." Some Board members voiced concern about the choice of the TLD string ".web" (See, Transcript, at 20:14 – 22:8), but believed that they could support the application if an alternative string were selected. See, Transcript, at 21:9-14. The final selection of a TLD string for Afilias, however, was not resolved at this time, and the Board moved on.

When the Board arrived at the Sarnoff application, a majority of the Board members present noted that they would approve it, with either the selected TLD string (".iii") or a different string:

TOUTON: Sarnoff, JV Team dot i i i. How many think that ought to be on the list? Six is all I get on that? Is that right?

DYSON: Would it be different if the string changes?

WILSON: [Nods yes]

MALE VOICE: Yes

TOUTON: Do people feel that having only six justifies changing the string?

VOICE: [Inaudible.]

TOUTON: Let's note that and go on.

See, Transcript, at 22:13 – 23:3.[5]

After reviewing all the applications, and leaving only selection of the TLD string for the Sarnoff and Afilias applications for final discussion, the Board returned to these two applications, taking the Sarnoff application first:

TOUTON: OK. Going back to Sarnoff/JVTeam i i i, where I think we had only six favoring, as it currently stands. Question was… is… would changing the string change that result? Anybody have any views? We are talking about Sarnoff/JVTeam's dot i i i proposal. It only got 6 votes in the straw poll. Let me ask those who did not vote for it, would that change your…? Linda. Would a different string change your mind?

WILSON: Yes, yes, yes. With a different string I would vote for it. I can make a suggestion for a string, but I don't think that's our job.

CERF: Well, ultimately it is our job to finally approve a string. We do have that respons...

WILSON: I'm supposing dot P E R S. It gets you out of the…

CERF: No, we are not doing that now, we are doing Sarnoff…

TOUTON: What she is saying is that…

CERF: Oh, she is talking about the other case. OK.

TOUTON: One way to handle this would be to say that they get one string and to require approval before execution of any agreement and then leave it for a matter for negotiation, expressing concern that that's not an appropriate string. Does that approach seem suitable to the Board?

WILSON: As long as it comes back to the Board for approval, I don't think it is a problem.

TOUTON: Yes, Ken?

FOCKLER: Is it possible to see because in their application they did state other strings so that we don't have to be seen to be making them up, but there might be something else there. Andrew mentioned some, but I'm sure in the application, that I don't have, and it wasn't listed in the book, but in the actual application…

VOICE: I am looking…

KRAAIJENBRINK: We might provisionally approve triple i, but in such negotiations with Sarnoff/JVTeam ask for a more appropriate string which is brought back to the Board for approval. I think that carries full support in the....

VOICE: I agree.

TOUTON: Give me just a moment to have the team find…

DYSON: It's here. It's here.

FOCKLER: Maybe I am wrong…

TOUTON: This was a, this was a...There was a little bit of difficulty there because on item E2 which was where they were supposed to list the string, they only listed i. But, in the text of their proposal, they listed others.

See, Transcript, at 23:15 – 26:4. In fact, this was incorrect. Section E2 of the Sarnoff Application specifically mentioned ".i," ".iii," ".idi," and ".one" as acceptable TLD strings. As Mr. Fockler found, in the context of the detailed overview in Section E2, several alternatives were proposed:

FOCKLER: "The adoption of alternate TLD strings such as dot i d i, i i i, or one, or some other string to be decided, will be considered although such string would be less distinctive."

DYSON: Dot one is available.

KRAAIJENBRINK: I am afraid we are entering a strange procedure here.

ROBERTS: I thought I might explain that I had a sort of generic objection here. We were specifically asked in view of the several ccTLDs that are already exploring this space not to do this at all. And I didn't think that was controlling. But I must say that I feel that we don't, aren't particularly under any obligation to do more than one to deal with the proof of concept at the worldwide TLD level.

TOUTON: Let me ask Joe, Joe Sims to make a few remarks.

SIMS: Just in the nature of general advice, what we have here is an application which has changed dramatically since the time it was made with this combination. The string is a string which is not acceptable, apparently, to a number of the members of the Board. They did not actually request additional strings and if you read the language in the application, it says they would consider additional strings. This is a series of reasons why at this particular time and this proof of concept stage, the Board might consider that this should be postponed to a later date.

TOUTON: Any reactions to that from the Board members?

WILSON: Sounds wise.

FOCKLER: Having heard that, could we have one more show of hands on triple i?

TOUTON: OK, Let's....do people support the triple i proposal being in the basket?

[two board members raise hands]

MALE VOICE: It went down, but it got noted.

TOUTON: OK. I think it should be taken out of the basket then, is my understanding? OK. Please take it out of the basket. We now have seven.

See, Transcript, at 26:6 – 28:5. At this point in the proceedings, the undersigned counsel attempted to contact Mr. Sims to explain, as detailed more completely in Section G(2) below, that the applications had not, in fact, changed at all and that the information he had just conveyed to the Board was wrong.

Although the proposal had been removed from consideration, Ms. Dyson sought further explanation for the reasons the Board had denied the Sarnoff application:

DYSON: Wait. Can we hear why they changed their minds? I thought this was a....

CERF: No, no. It was the combining of two proposals together in the last minute...to create a joint.

DYSON: Yeah, no, I know. No, no, not why they changed their mind, why we changed our mind.

CERF: I will be happy to answer that. i i i does not...I'm like Mike. It doesn't ring beautifully on my brain. So, I would love to find something else. But, I would rather that the applicant find it rather than we make it up for them.

DYSON: Yeah. But, if we do that... As I understood, the proposal was to go back to them and have them in negotiation with us to come up with a new string, where it is their string. But, we instruct the Staff to go forward with the proposal to amend it to suit us.

CERF: Well, as I understood Joe Sims' observation. Since the proposal is no longer what either of them were, it is a combined thing which has not been submitted to us as a combined proposal. It has simply been orally conveyed to us that they are jointly willing to work together. I understood Joe's comment to be a caution that we haven't actually seen a proposal of the joint group. We have only seen two different proposals and an oral commitment to work together. Is that a fair....

See, Transcript, at 28:7 – 29:12. Having now had a brief, hurried conversation with the undersigned counsel, Mr. Sims tried to correct his earlier mistake, though the retraction was incomplete.

SIMS: Yes, let me just add one point just to, just to make sure there is no confusion here. The applicants have said that they have left their individual applications in the...pending...so that there are two separate applications here. They have joined themselves together in the terms of a... of a... to let you know that they have joined together for technical purposes. So, I don't want there to be confusion on this. My point was that when we are trying to find a relatively limited number of very attractive applications to the extent there are significant questions or uncertainties about a particular application, that is a factor that the Board could, and in my view should, take into account in deciding whether to proceed.

TOUTON: Let me just make clear. There was a one-page sheet that was given to us yesterday... I don't recall whether it was distributed to the Board... describing in obviously very general terms how they would combine these two very extensive proposals. Vint.

CERF: In that case, I would like to argue in favor of keeping them on the list, but having them in negotiation that produces a string that we approve.

[The audience claps]

TOUTON: Please... please no audience reaction. Again, you are here to observe, not to participate. OK.

KRAIINJENBRINK: Do we need a third vote?

TOUTON: Ok, let's do another straw...

KRAAIJENBRINK: I'm asking because we made another vote on the proposal and that vote was only two for. The first vote was only six for out of the eleven or twelve. So, I know that in America the recounting is a.....

[Laughter from Board and Audience]

TOUTON: Alright. In a moment, I will have another straw poll. Is there any more commentary about it though first?

CERF: I feel a need to respond to Hans' observation. That is, we did not vote on the proposal I just put on the table specifically, which is to remand the string decision to negotiation and I think we should vote on this.

KRAAIJENBRINK: I believe we should vote on every proposal in a formal way, maybe, if necessary. It was straw poll. But going from straw poll to straw poll..... If you need a third straw poll, then do it.

CERF: We can remand this to the State of Florida at some point.

TOUTON: I would like to avoid having more than three straw polls. Are there any other comments that anybody wants to make about this....it's not up there right now....Sarnoff JVTeam proposal. Thank you, Ben. OK. There being no comments, let me do a straw poll. First... I think the first straw poll established... well, I am not sure what it established. Which Board members believe that Sarnoff JVTeam proposal should be on the list with a requirement that the dot i i i string be changed to another string to be negotiated and then approved by the Board? Is that the right straw poll, first of all?

CERF: Yes.

TOUTON: Which think that? One, two. Oh wait, there is a point of clarification

FOCKLER: Does that in any way impact any other decision that we have made today? I don't think so for any of the other applications...

TOUTON: Andrew, you have some sage advice.

McLAUGHLIN: Jun will clobber me for saying this. The only one that I suppose that affects is that dot geo was taken out in part, I think, because it was the third instantiation of a registry by the same people. So, just so that the record is clear, that was one of the factors you talked about when taking out dot geo.

FOCKLER: Was it focused just on the string?

McLAUGHLIN: No. No.

TOUTON: Ken was asking a broader question. Now, let's take the straw poll. Straw poll is -- should Sarnoff/JVTeam be put on the list with the understanding dot i i i is not acceptable, that a new string will be negotiated and will be approved only when approved by the Board? OK, another question.

WILSON: My question is under our procedures, is taking that step well within our... I am asking whether given the announced procedures, is taking this step well within....an acceptable range? Could we have some advice about that?

[Pause]

SIMS: I don't think it would be correct to give you a definitive yes or no answer on that, for which I apologize. But I think it would be fair to say that in the context of this entire process, to undertake a negotiation of this type with one of the applicants would raise some potential for claims of unfairness or discriminatory treatment. Now, I'm not prepared to say that those claims could not be dealt with effectively if they were raised, but I also think that it would raise the potential for that sort of claim to be made.

TOUTON: Let me just amplify on that. I do think that's right and I think that actually involves not only the string but also the joinder of the two applications. Other applicants may complain that they weren't permitted to change their application in a substantial way after submitting it. Alright. Let's do the poll I was doing on Sarnoff JVTeam as to whether people thought that that proposal ought to be on the list with the understanding that dot i i i would not be the string but that it would be negotiated, subject to final approval by the Board. How many feel that way?

[No votes] [Laughter from the Board]

CERF: It's all in how you ask the question, isn't it?

TOUTON: We should offer our services to the West Palm Beach....

DYSON: Can I ask for a different ballot?

TOUTON: Alright. Please take that one off the list.

Transcript, at p. 29:14 – 35:2.

2. ICANN's Counsel Presented the Board With Incorrect Information About the Nature of the Sarnoff-NeuLevel Joint Venture, and Such Information Was Used to Sarnoff's Detriment When the Vote Was Called on its Application.

As Mr. Cerf noted after the final vote on the Sarnoff application, "It's all in how you ask the question." Transcript, at 34:20. In Sarnoff's case, "the question" was called for a vote against the background that the application had been changed in a "substantial way" and that entering a negotiation with an applicant on a more aesthetically pleasing TLD string might be unfair or discriminatory. The first concern was misplaced, and the second should never have been asked. At the end of the day, the only "unfairness or discriminatory treatment" came in the handling of the Sarnoff application relative to other applications.

As the exchanges noted above make clear, not once in the first two application reviews did anyone raise concerns about the Sarnoff application, the partnership of the talents of Sarnoff and NeuLevel, or the requested TLD strings. But the tone of the discussions changed dramatically after ICANN's outside counsel, Joe Sims, introduced the following concern:

Just in the nature of general advice, what we have here is an application that has changed dramatically since the time it was made with this combination. The string is a string which is not acceptable, apparently, to a number of the members of the board. They did not actually request additional strings and if you read the language in the application, it says they would consider additional strings.[6] This is a series of reasons why at this particular time and this proof of concept stage, the Board might consider this should be postponed to a later date.

See, Transcript, at 27:2-12. Importantly, the idea that the application had "changed dramatically since the time it was made was objectively incorrect.

Prior to the Board meeting, Sarnoff and NeuLevel circulated a one-page overview describing the nature of their joint venture. See, "Overview of Sarnoff, Inc. and NeuLevel, LLC Agreement" (copy attached as Exhibit "B"). As this document made clear, Sarnoff and NeuLevel had simply "assign[ed] all rights and interests in their respective ICANN top-level domain applications to their new partnership." The applications were neither withdrawn nor merged. The only change made to the applications was the request that the TLD string requested by each application be amended "so that each now requests delegation of the new top-level domain ‘.i,' [including the fallback presented to the Board, ".iii"] ‘.nom,' or ‘.per' (in order of preference)." Id. The sole point of the amendment was to state that the new partnership was seeking common TLD strings for their two applications and to make clear that the parties were not seeking delegation of two separate TLDs.

Sarnoff and NeuLevel were required to describe their new partnership in the one- page overview because, in a surprise departure from the established protocol for the Public Forum presentations, they were permitted only three minutes to make their presentation for both applications, rather than three minutes each. Prior to the public forum on Wednesday, November 15, 2000, the undersigned counsel asked ICANN staff about the process issues involved in informing the Board of the assignment of the applications and the agreement by Sarnoff and NeuLevel to work together in partnership. During that discussion with staff, counsel was apprised that Sarnoff and NeuLevel would be permitted six minutes to make their presentation, as they had two applications before the Board. As promised, ICANN staff worked together with the technical support provided by the Berkman Center to set the time clock for Sarnoff and NeuLevel (which coincidentally were back-to-back on the program) collectively at six minutes. Unfortunately, and without waiting to hear why the clock had been set at six minutes or any other details of the two applicants' arrangement, Ms. Dyson, acting as Chair, required that the clock be set at only three minutes for Sarnoff and NeuLevel together. See, Transcript, at 2:3-8. As a consequence, Shailendra Suman, Executive Director of Sarnoff Corporation and the person designated to speak about both the Sarnoff and NeuLevel applications, was required to omit important segments of his presentation. This departure from established procedure certainly contributed to the Board, its staff, and its counsel having an incorrect understanding of the two applications.

Nevertheless, to make its position clear, Sarnoff submitted the one-page overview later in the afternoon as a handout which indicated that this was merely an assignment of interests, not a merger of applications. By assigning their rights in the respective applications to their partnership, Sarnoff and NeuLevel were agreeing to work with each other in the context of whichever application was selected and approved by ICANN. They did not merge or "dramatically change" the applications precisely for the reasons apparently underlying Mr. Sims' concern: that the Board and Staff would be unable to review a substantially changed application in the limited time allotted. The one-page overview made this clear, and it directly contradicts the contention made by counsel that the applications had "changed dramatically."

The parties believed that the skill sets of the two companies were complementary and that, working together, they could better fulfill the business and technical plans set out in either application. Members of the Board apparently agreed.[7] This was simply an instance in which each application was enriched by bringing the complimentary talents of a new party to bear in its execution; it was not a change of naming policies, business plans, or technical specifications.

As noted, supra, following the presentation of this erroneous information, the undersigned counsel asked for Mr. Sims' attention, and in a side bar conversation explained that the information just delivered to the Board was incorrect. After that conversation, Mr. Sims attempted to correct his mistake -- a correction that prompted at least one Board member to ask for the Sarnoff application to be accepted.[8] Nevertheless, in spite of Mr. Sims' correction, the factual mistake was repeated by Mr. Touton when the Sarnoff proposal was finally called for a vote:

SIMS: … I think it would be fair to say that in the context of this entire process, to undertake a negotiation [for an acceptable TLD string] of this type with one of the applicants would raise some potential for unfairness or discriminatory treatment. Now, I am not prepared to say that those claims could not be dealt with effectively if they were raised, but I do think that it would raise the potential for that sort of claim to be made.

TOUTON: Let me just amplify on that. I do think that's right and I think that actually involves not only the string but also the joinder of the two applications. Other applicants may complain that they weren't permitted to change their application in a substantial way after submitting it. Alright. The poll I was doing on Sarnoff/JVTeam is to whether people thought that that proposal ought to be on the list with the understanding that a new string would be negotiated, subject to final approval by the board. How many of you feel that way?

[No votes] [Laughter from the Board]

CERF: It's all in how you ask the question, isn't it?

TOUTON: We should offer our services to the West Palm Beach....

DYSON: Can I ask for a different ballot?

TOUTON: Alright. Please take that one off the list.

Transcript, at 33:22 – 35:2 (emphasis added).[9]

In fact, the Sarnoff and NeuLevel applications had not changed, and including this important misstatement of fact in the very question presented to and relied on by the Board unfairly tainted the vote. Had the Board not been presented with this incorrect information, Sarnoff believes that its application would have been included in Board Resolution 00.89.

3. The Published Evaluation Criteria for New TLD Applications Did Not Contain Any Requirement that the Chosen TLD String Be Pronounceable, Mnemonic, Or Aesthetically Pleasing.

In its "Criteria for Assessing TLD Proposals" and its responses to questions that were publicly posted on the ICANN web site, ICANN set minimal restrictions on an applicant's choice of a TLD string. It noted a handful of absolute prohibitions[10] and made a few additional cautionary advisories for the selection of appropriate strings,[11] but it generally appeared to give applicants wide discretion in selecting a string.

The "appropriateness" of the TLD string was only to be a factor in considering whether the TLD would "enhance the utility of the DNS." See, Category 4, "Criteria for Assessing TLD Proposals." Specifically, ICANN noted the following:

One motivation often cited for introducing new TLDs is that doing so might increase the utility of the DNS. Under this view, the appropriateness of adding new TLDs should be evaluated based on whether addition of the new TLDs:

would sensibly add to the existing DNS hierarchy and

would not create or add to confusion of Internet users in locating the Internet resources they seek.

At least the following considerations will be considered in this regard:

a. If the TLD is intended for a particular use or purpose, does the TLD label suggest that use? Is this true for a large portion of Internet users globally (i.e. in different languages)?

b. Is the proposed TLD semantically "far" from existing TLDs, so that confusion is avoided? (For example, TLD labels suggesting similar meanings might be more easily confused.) Is it phonetically distinct from existing TLDs? Meanings and pronunciations in different languages may be relevant to these inquiries.[12]

These provisions are quoted in full because, in context, it is clear that the primary purpose for evaluating a TLD string is to ensure that users are not confused. The TLD string selected should not be easily confused with another TLD, nor should it create user expectations that are contrary to the actual registrations found within the TLD.

Subject to those concerns, however, TLD applicants presumably would have had great latitude in the selection of their TLD string. Certainly there was no requirement that the TLD be pronounceable, mnemonic, or sound beautiful. While not conceding that any of those criticisms are true,[13] neither ".i" nor ".iii" violated any of the written policy considerations noted above.

ICANN had stated that single letter TLDs were reserved for "future expansion" of the DNS.[14] Fully cognizant of the possible limitation on the delegation of a single-letter TLD, Sarnoff had proposed several alternatives, including ".iii," ".idi," and ".one." NeuLevel, in its application, had proposed ".per" and ".nom." In its one-page overview of the new joint venture, Sarnoff prioritized the TLD selections for the convenience of the Board as ".i," (including the three-digit fallback presented to the Board at the Public Forum, ".iii"), ".nom," and ".per."

In spite of the wide number of TLDs proposed, any of which would have been acceptable to Sarnoff, the primary TLDs of ".i" and ".iii" satisfied all of the published criteria on TLD strings. They were semantically far from existing TLDs and the use of the character "i," either by itself or as a triptych, is easily branded as a space for individuals.

Even if Mr. Roberts' concern that the chosen TLD string said "nothing about what that might actually mean" was reasonably within the umbra of the TLD evaluation criteria, it misses the point behind Sarnoff's selection of that string and ignores the substantial marketing and branding efforts that Sarnoff had pledged to put behind creating a unique identity for the TLD.[15] Sarnoff specifically selected these TLD strings because they could appeal equally to users around the world and not have a specific mnemonic connotation with any single language. This was specified in the application:

As a personal identifier, addresses within the TLD are expected to find a variety of applications that are not readily characterizable as being "commercial" (as in the original charter for .com), "network oriented" (as in the original charter for .net), or "organizational" (as in the original charter for .org). Rather than limit the perceived scope of applicability of addresses within the TLD by selecting a multiple character string, which might suggest a particular application in one or more languages, ".i" is relatively free of cognitive associations other than perhaps "Internet", "identity", "individual", or the first-person pronoun, "I".

See, Sarnoff's "Description of TLD Policies," at E2.[16]

The primary purpose of the TLD was to service individuals around the world, not simply those in any single part of it. The requested TLD strings, ".i" and its three-digit counterpart ".iii," were selected because of their loose, but direct, association with many words and concepts, both in English and other languages, relevant to the Sarnoff business plan (in English, words such as ‘individual,' ‘internet,' ‘identifier,' and ‘I').[17]

The application provided an overview of the significant branding efforts, resources and personnel that Sarnoff, its business consultant partner Atomic Tangerine, and its retained public relations firm Ruder Finn, planned to undertake in all parts of the world to equate the TLD in the minds of users everywhere as a space for individuals. By intentionally selecting a TLD string that was not a shortened version of a particular word in a particular language, Sarnoff had the flexibility to market the TLD worldwide. That ".iii" is, at present, unknown to users is not surprising as the TLD does not yet exist. The more relevant point is that Sarnoff had outlined a path to brand that particular TLD string as a name space for an individual's Internet identity.

Regardless of the merits of the TLD string in Sarnoff's or the Board's judgment, the fact is clear that ".iii" did not violate any naming prohibition, it did not conflict with any existing TLD, nor was it easily confused with any existing TLD. Accordingly, any non-offensive selection within those boundaries was Sarnoff's to make, and the Board should have given deference to the applicant's selection.[18] Had the Board judged the string presented to it at the Public Forum on Wednesday, November 15, 2000 -- ".iii" -- strictly against the published evaluation criteria, Sarnoff believes that its application would have been included in Board Resolution 00.89.

4. The Board Had Discretion to Approve Sarnoff's Application, Subject to Additional Negotiation on An Appropriate TLD String.

As noted above, Sarnoff believes that the better reading of the "Criteria for Evaluating TLD Proposals," published by the Board in advance of its Annual Meeting, gives great discretion to the applicant in the selection of a TLD string. Nevertheless, if the Board reads those criteria to require a finer, more subjective analysis by the Board of the TLD string, including an assessment of its aesthetic qualities, Sarnoff believes there was no prohibition against approving its application, subject only to future, final agreement on an appropriate TLD string. This course of action would have been especially appropriate in Sarnoff's case, given the fact that the Board had diverged from the published evaluation criteria by focusing on the aesthetic qualities of Sarnoff's string selections.[19]

During the discussion of the Sarnoff application, several Board members recommended that it be approved, subject only to final approval of an acceptable string. The sole reason that this course of action was not followed appears to have been due to a fear that such a procedure might be unfair to other applicants. This concern, however, fails to appreciate the advanced stage of the discussions when the question of whether the Sarnoff TLD string was acceptable arose. By that time in the review process, over three-quarters of the applications had fallen from consideration because of technical issues, naming conventions for SLDs, lack of support from the relevant community, or the failure of some other item detailed in the "Criteria for Evaluating TLD Proposals." At the time that the Sarnoff application was finally denied, it was one of only eight applications still before the Board for approval; all of the other 39 applications had been declined for other reasons. Importantly, to the best of Petitioner's knowledge and recollection, none of the other 39 applications declined were taken out of consideration solely on the basis of the selected TLD string.

The only unfairness to an applicant through not considering alternative TLD strings was in the treatment of Sarnoff. Of the final eight applications before the Board, the Board discussed the selected TLD string for three of them: ".air," ".web," and ".iii." In each of the first two instances, the Board picked an alternative TLD string from the applicant's application, ".aero" for ".air" and ".info" for ".web."[20] Having reached into the application for alternative strings for these two successful applicants, the Board should have done the same for Sarnoff.[21] Given that the Sarnoff application was rejected on the basis of a criterion not presented to the applicants during the application stage, the Board should have allowed a negotiation of an appropriate TLD string.

5. Protecting ccTLD Interests in Creating Country-Specific Personal Name Spaces Was an Inappropriate Concern That, Once Raised, Unfairly Biased the Vote.

Though it is not specified in documents publicly available from the ICANN web site, some ccTLDs apparently had asked ICANN not to accept any applications for personal domain names.[22] As Mr. Roberts stated during the discussion of the Sarnoff application:

ROBERTS: I thought I might explain that I had a sort of generic objection here. We were specifically asked in view of the several ccTLDs that are already exploring this space not to do this at all. And I didn't think that was controlling. But I must say that I feel that we aren't under any particular obligation to do more than one to deal with the proof of concept at the worldwide TLD level.

See, Transcript, at 26:16-23. Not only was this consideration "not controlling," as Mr. Roberts acknowledged, but it was not relevant and was, quite possibly, prejudicial to the Board's consideration.

Individual name spaces in a handful of country-specific TLDs are no solution to individuals in other parts of the world. Limiting competition in the personal domain name space to one new gTLD – ".name" – solely for the protection of the handful of ccTLDs interested in exploring this name space, is counter to ICANN's mandate to foster competition and hurts the majority of users worldwide whose ccTLDs are not exploring this space.[23] This factor should not have been introduced into the decision-making process, as it was not only outside the bounds of the published evaluation criteria, but directly contrary to it.[24]

Importantly, the Sarnoff application proposed to make available free personal domain names to anyone, anywhere in the world:

Free Personal Numeric Domain Names: Domain names of the form tendigits.twoletters.TLD (as described in D13.2.1 above) will be provided free of charge in order to provide equitable access to a Personal Domain Name irrespective of an individual's ability to pay. Our projections predict that in Year 1 around 20% of registrations in the proposed TLD will be requests for free numeric domain names, a figure predicted to grow to 40% by the end of Year 4. (See Section D13.2.5, Estimated Demand for Registry Services, for further information.) Revenues from fee-based registrations will be used in part to subsidize the free domain name program.

See, Sarnoff's "Registry Operator Proposal," Section D.13.2.2. As detailed in the application, these free domain names would provide deep penetration of the DNS into new, underserved communities. Of course existing ccTLD registries asked ICANN "not to do this at all;" the Sarnoff application was a direct competitive threat.[25]

While Mr. Roberts correctly explained that the ccTLD's protectionist concerns were not controlling,[26] the introduction of such a concern, which stood in direct contrast to the published evaluation criteria, was inappropriate and may have prejudiced the outcome of the Board's deliberations.

* * * * *

III. CONCLUSION.

Sarnoff appreciates the fact that the Board was measuring TLD applications against complex and varied evaluation criteria. Given the open, free-form discussion that characterized the Board's evaluation process, some departure from the published evaluation criteria was inevitable. Nevertheless, when plain factual errors and factors outside the published evaluation criteria are determinative of an application's fate, as here, that decision should be corrected. The open and transparent nature of the Board's Annual Meeting makes it possible to look back on the public record made during the Board meeting, measure the decisions that were reached at that meeting against the published evaluation criteria, and, when mistakes were made, use the RFR process to correct them. Indeed, part of the "proof of concept" for the TLD evaluation process necessarily includes using the RFR process, when invoked, to ensure that mistakes can be corrected.

Adding the Sarnoff proposal to Board Resolution 00.89 also would enhance the testbed, as it would provide the Board a rare opportunity to launch competitive registries targeting similar markets at the same time and on an equal, neutral basis. This would allow the review process to consider the effects, if any, of competition on the initial launch and early operation of new TLDs.

Sarnoff submits that the rejection of its application was based on the Board's reliance on an incorrect statement of the facts, a departure from the published TLD evaluation criteria, and a consequent disparate treatment accorded to Sarnoff in the selection of an alternate string relative to the treatment of .aero and .info. For the reasons stated above, Sarnoff respectfully requests that the Reconsideration Committee recommend, and the Board as a whole agree, that the Sarnoff application for ".iii" (or another string to be agreed) be included in an appropriate amendment to Board Resolution 00.89.

Respectfully submitted,

SARNOFF CORPORATION

By its attorneys,

Bret A. Fausett
HANCOCK ROTHERT & BUNSHOFT LLP
515 South Figueroa Street, 16^th Floor
Los Angeles, California 90071-3334
Tel: (213) 623-7777
Fax: (213) 623-5405
E-mail: baf@fausett.com

[1] Sarnoff also requests that, if the Board as a whole reconsiders the Sarnoff application at a non-public telephonic Board meeting that either a recording be made of such meeting or that the undersigned counsel be permitted at least a "listen only" line to the Board meeting for the purpose of recording the call and preserving the record for any subsequent Independent Review.

[2] In preparing this RFR, Sarnoff has transcribed relevant sections of the Board's meeting as available on the RealVideo created and hosted by Harvard University's Berkman Center for Internet and Society. A copy of the transcribed sections is attached to this RFR as "Exhibit A." To the extent that the Reconsideration Committee, or the Board as a whole, believes that it needs to review additional transcribed sections of the RealVideo in order to decide the matters presented by this RFR, Sarnoff would be pleased to provide another transcript upon request.

[3] The favorable Staff Review and the lack of concern stated by Board members about technical and naming policies is important to note because, as stated in the published "Criteria for Assessing TLD Proposals": "ICANN's first priority is to preserve the stability of the Internet, including the domain-name system (DNS)." See, Criteria for Assessing TLD Proposals, http://www.icann.org/tlds/tld-criteria-15aug00.htm. There is no indication anywhere in the published record that the Sarnoff application posed any threat to the stability of the Internet or the DNS. For example, as noted in the ICANN Staff Report, the Sarnoff proposal paid "particularly strong attention to the reliability and security aspects of running the registry." Staff also noted that the proposal was "a model for registry security." See, "Staff Summary of Sarnoff Application," at http://www.icann.org/tlds/report/i1.html.

[4] Using the "shopping cart" analogy of the day, Mr. Kraaijenbrink stated: "I believe we have the shopping cart here with eight very viable proposals, with a spread of TLD strings which cover diverse applications as we have evaluated them from the proposals. So, I would like to find my credit card and go to the cash register." See, Transcript, at 17:2-7.

[5] It is worth noting that, in this straw poll, the Sarnoff application received six votes in its favor based on the TLD string ".iii," with two more Board members stating that they would vote in favor of the application with a different TLD string. Collectively, these eight votes were equal to the number of votes garnered by other successful applicants for their prospective TLD delegations. Even with the six votes alone, there was a possibility, under the rules being used by the Board, that consensus had been reached. As Mr. Sims noted later in the day with regard to the award of ".web" to Afilias: "OK, so we have six in favor, three against and the rest abstaining. So the Board must use its own judgment to decide whether that is the consensus." See, Transcript, at 41:20-23.

[6] The idea that Sarnoff had "not actually request[ed] additional strings" but merely said that it would "consider" them was incorrect, as both the presentation at the Public Forum (at which Sarnoff requested ".iii") and the one-page overview provided to the Board revealed. The Application itself also made this clear. The very narrow reading given to Section E2 of the Application by Mr. Sims is simply not a reasonable one given Sarnoff's obvious understanding that the primary TLD string requested, ".i," might be held in reservation by ICANN and the statements made by Mr. Suman to the Board at the Public Forum.

[7] As Mr. Cerf commented: "They [Sarnoff and NeuLevel] represent a strong proposal, especially in combination." See, Transcript, at 8:8-9 (emphasis added).

[8] Upon learning that Sarnoff and NeuLevel had, in fact, made clear the nature of the proposal that was before the Board, Mr. Cerf stated: "In that case, I would like to argue in favor of keeping them on the list, but having them in a negotiation that produces a string that we approve." See, Transcript, at 30:10-12.

[9] Even if the Sarnoff and NeuLevel applications had been merged, the idea that other applicants might object appears more conjecture than reality. At the conclusion of the Public Forum, there was ample opportunity for other applicants to object. To the best of Sarnoff's knowledge, none did.

[10] Specifically, ICANN wrote that (1) TLD strings must contain only the letters A to Z, digits, and hyphens, all coded to the rules of the "ASCII" character set. See, FAQ No. 5, ICANN "TLD Application Process FAQs," http://www.icann.org/en/tlds/tld-faqs.htm. (hereafter, "TLD FAQ"); (2) TLD strings may not start or end with a hyphen, start with a digit, or exceed 63 characters in length. See, TLD FAQ, Nos. 5 and 36; (3) TLD strings should not match alpha-2 codes on the ISO 3166-1 list. See, TLD FAQ, Nos. 17, 21, and 24; and, (4) no TLD string should consist of two-letters unless the ISO 3166 Maintenance Agency has indicated that the proposed two-letter code will not be placed on the ISO 3166-1 list in such a way that would be incompatible with the proposal. See, TLD FAQ, No. 47.

[11] TLD strings should be semantically "far" from existing TLD strings, so that user confusion is avoided. See, TLD FAQ, No. 24. TLD strings should generally be between three and sixty-three characters long. See, TLD FAQ, No. 47. And ICANN noted that, under current practice, one-letter codes were reserved for future DNS extensibility. See, TLD FAQ, No. 47.

[12] The TLD Criteria also included two additional notes on the selection of TLD strings, though neither are applicable here: Whether the proposed TLD avoids names reserved by existing RFCs (See, TLD Criteria, 4(c)); and, whether, in the case of restricted TLDs, the restriction is one that will assist users in locating domain names. (See, TLD Criteria, 4(d)).

[13] While the evaluation process was not designed to assess the aesthetic beauty of the TLD string, Sarnoff respectfully disagrees with the conclusion of some Board members that .iii was not pronounceable ("dot triple i"), not mnemonic ("individual," "Internet," "I," and "identity" are all concepts that Sarnoff planned to tie to the ".iii" brand), or did not ring beautifully in the brain (there's poetry in "dotcom"?).

[14] In its application, Sarnoff believed that it had made a compelling case for the delegation of the single-letter ".i," but unfortunately, there was no discussion among the members of the Board as to whether the creation of a personal name space for individuals was a sufficient "future expansion" to delegate that particular requested TLD string.

[15] The Staff Report specifically credited the marketing focus of the Sarnoff application: "The strengths of this application include a thorough assessment of the market by target segment. It contains a strong assessment of the marketing mix including a plan by quarter." See "Staff Summary of Sarnoff Application," supra, at C.2.

[16] The overview of the TLD selection continued: "For the proposed target applications, other TLD strings may be appropriate. However, ".i" is less likely to be confused with .id (Indonesia) or .int (International Organizations) than other logical English mnemonics for terms such as "Internet", "identity" or "individual".  It is noted that single letter TLDs have been reserved for unspecified future expansion of the Internet domain name space, and it is not clear what "expansion" is intended. For example, it would make little sense to introduce a ".com.i" in order to provide for additional .com addresses, since such use would lead to obvious confusion. The new paradigm embodied in the .i TLD proposal of assigning domain names for individual personal use as well as Internet enabled devices is the sort of application for which single letter domain names are reserved. Because the .i TLD is targeted for use by individuals, an individual TLD character reinforces the perception that the TLD is not intended as a supplemental generic TLD. Adoption of an alternative TLD string such as ".idi" (Internet Digital Identifier) or .iii, or .one, or some other string to be decided would be considered, although such a string would be less distinctive and therefore pose a greater possibility of being confused with a traditional TLD."

[17] For example, the application highlights Mr. Cerf's maxim of "The Internet is for Everyone," which, in the context of individual users targeted by the application, translates to the Internet is for individuals, or "iii."

[18] Mr. Kraaijenbrink appeared to appreciate that the Board's had departed from its published evaluation criteria when he commented: "I believe we are entering a strange procedure here." See, Transcript, at 26:13-14 (emphasis added). This insightful comment came on the tail-end of the Board's speculation on possible substitute strings for the Sarnoff application.

[19] The Board briefly considered the string ".per," one of the strings specifically requested in the one-page overview, but ultimately decided that it might conflict with the previously unreserved three-letter ISO-3166-1 code for the nation of Peru. See, Transcript, at 12:13-16. Ironically, under the same rules, ".com" would conflict with the three-letter code, COM, given to the Islamic Federal Republic of the Comoros.

[20] The concerns about the Sarnoff application were much similar to the concerns about the ".air" application. There was no concern "with the business proposition or the technical viability," but only with the choice of the TLD string itself. See, Transcript, at 9:24 – 10:1. For himself, Mr. Cerf clearly believed that such concerns were insufficient to delete an application from consideration. As he stated with regard to ".air": "I am reluctant to not put it in the basket because it is a reasonable proposal except for the concern that's been expressed about the generic word." Transcript, at 10:16-19. The same considerations underlying the change from ".air" to ".aero" should have allowed a final negotiation with Sarnoff.

[21] As noted, infra, Sarnoff proposed the strings ".i," ".iii," ‘.idi" and ".one," and NeuLevel proposed the strings ".per" and ".nom."

[22] The Netherlands, for example, timed the opening of .nl to individuals for November 15, 2000, the day that the new TLD applications were presented to the Board and the Internet community at the ICANN public forum. See, ".nl Opens for Individuals Today," DomainInfo.com, November 15, 2000, at http://domaininfo.com/news.asp?id=263&lang=en. One "open" ccTLD has even deceptively offered to provide "free" personal domain names -- subject only to a $4.95 "verification fee." See, "Welcome to Hot.MU – domain name registration services," at http://www.hot.mu/personalname.html. (The official ccTLD registry for .mu also provides the service "Hot Babes" at http://www.hot.mu/hotbabes.html.) If this represents the kind of individual name space offered by the ccTLDs, then clearly competition is long overdue.

[23] The Staff Report seemed to appreciate this concern, as it specifically called for the selection of "a limited number" of the applications for "personal name" TLDs. See, "Report on TLD Applications," Section III.B(1)(b), "Personal Group" (emphasis added).

[24] "As noted in the White Paper, market mechanisms that support competition and consumer choice should, where possible, drive the management of the DNS. One of ICANN's core principles is the encouragement of competition at both the registry and registrar levels." See, "Criteria for Assessing TLD Proposals," at http://www.icann.org/tlds/tld-criteria-15aug00.htm

[25] Fostering competition is not only one of the stated evaluation criteria, but it is also a fundamental tenet of ICANN's Articles of Incorporation: "The corporation shall operate for the benefit of the Internet community as a whole, . . . through open and transparent processes that enable competition and open entry in Internet-related markets." See, Articles of Incorporation, as revised November 21, 1998, at Article 4.

[26] In spite of the fact that some ccTLD administrators apparently advocated against new personal name space, the Governmental Advisory Committee counseled that "ICANN should be particularly attentive to ensuring that adequate competition is introduced at all possible levels in the DNS registration market." See, "Opinion of the ICANN-GAC on New Generic Top Level Domains," 16 November 2000, at 1.3. Presumably, introducing competition "at all levels" would include introducing competitive registries in the personal domain name market.

Exhibit A

Exhibit B

Page Updated 22-Sep-2009
©2002 The Internet Corporation for Assigned Names and Numbers. All rights reserved.