July 16, 2000 - Yokohama, Japan
I. Welcome - Dyson
A. Meeting open to observation,
II. Agenda - Roberts
A. Proposed Top-Level Domains
B. At Large Elections
C. Root Servers
D. Other Matters
E. Cerf: Suggest moving forward
and making additions later.
III. Top Level Domains -
posted – based on DNSO WG, NC process, NC recommendations, public comments
received, public forum. Resolves that new TLDs be introduced in a measured
and responsible manner.
B. Schedule (to be adjusted if
1. Proposals requested
2. Deadline for
receipt of applications (October 1)
3. Close of public
comment period (October 15)
4. ICANN announces
selections (November 20)
5. Target for completion
of negotiations (Dec 1)
C. Fees – non-refundable $50,000
per application to become a sponsor or operator of a registry
D. President should seek information
1. Technical, business,
and management plans
2. Policies to promote
orderly registration during introduction of TLD
3. Protection for
consumers in case of registry failure
4. Minimizing use
of TLD for infringements of intellectual property rights
E. Choice should consider:
1. Protections for
2. Extent to which
it leads to “proof of concept”
3. Enhancement of
competition of registration services
4. Enhancement of
utility of DNS
5. Evaluation of
delegation of policy functions for special-purpose TLDs
6. Whether proposal
meets previously-unmet needs
1. Wong: Ordering
of principles implies priority of importance?
• Touton: No. Other
principles may also be important.
2. Cerf: Who decides
what is to be published? Applicant?
• Touton: Intend
not to specify it in the language here. Follow procedure used for registrar
accreditation. Have to discuss details with particular applicants.
• Dyson: Let’s give
the President guidance that this should be an open process. If anyone’s
idea is worth something only because they kept it a secret, maybe the
idea is not worth so much in the first place.
• Touton: Procedure
for registrar accreditation: Anyone who thinks part of application is
confidential says so with application. Staff reviews. Either registrar
concurs with staff recommendation re whether or not confidential, or
removes that portion of application.
3. Wilson: Policy
is too vague? Language is not clear.
• Touton: “Measured
and responsible” per Names Council’s recommendation. Intend the wording
to reflect that this is a new endeavor.
4. Triana: Consider
financial capabilities also?
• Touton: Yes, add
5. Abril i Abril:
Should say “protect users” not “protect consumers” because more general?
• Touton: “Consumers”
is not the right word. OK to make the change to “users”?
6. Dyson: Costs
section should reflect costs of “launch” also?
7. Cohen: There
are other kinds of intellectual property abuses than infringements.
• Change made.
8. Fockler: Note
that ASOs and PSOs had the opportunity to review also, and no negative
comment received from either. Add additional “Whereas” clauses.
Specify some particular kinds of TLDs that might be added: Fully open,
restricted/chartered with limited scope, non-commercial, personal.
10. Wilson: Troubled
by delegation of policy formulation. Don’t want to delegate too much
from the Board to the President and staff.
• Touton: Want to
retain flexibility to design a proposal – but with the Board’s approval.
• Dyson: Want diversity.
Try delegating policy for some domains some way, some domains another
way. Need diversity on all fronts, including geography.
• Roberts: Note
that President is not authorized to execute an agreement with a registry
until the Board approves.
• Touton: Note that
schedule has been adjusted to accommodate the tentative Board meeting
schedule. Make this clear by adding “After approval by the Board” to
November 20 line.
11. Dyson: Add “variety
of business models and variety of geographic locations” to diversity
• Cerf: Can’t list
all dimensions of variation. And can’t exercise all dimensions of variation.
12. Fockler: Would
like to help President in the evaluation process?
• Roberts: President
is authorized to seek technical advice as appropriate.
13. Wong: Specify
14. Cerf: Expect
to receive proposals in English?
• Touton: That was
the anticipation. Could specify as much in the resolution. Or could
leave it to the President’s discretion?
• Roberts: Look
forward to a multilingual environment. .COM registration applications
have come from all over the world, and English requirement does not
seem to have presented a significant burden.
15. Cerf: Develop
a Q&A document? Web only? How about an email autoresponder?
16. Dyson: Anticipated
number? What if not enough good applications? Hope for ~10 applications,
• Touton: Staff
thinks it’s important to be careful in introduction of new TLDs. Ten
would be a significant strain on our ability to perform.
17. Abril i Abril:
Frustrated about low quality of debate. Registrars, ccTLD registries,
etc. all agree. Being unclear about what we expect raises the cost of
participation by applicants and others.
Hope to select one to two good proposals in each category. Add from
three to eight at the most, probably really more like six at most.
19. Dyson: Want
to be clear that we’re working on diversity. Want to prove that market
regulation will work – put in place rules to make sure the market works
(i.e. Whois, escrow, etc.). Willing to raise application charge if we
20. Fockler: Support
Dyson’s diversity comments. Hope to get to the upper part of the six-to-eight-to-ten
range. But disagree that it’s a negative factor for proposers to want
to keep aspects of their applications private.
21. Cerf: This is
a proof of concept. Let’s not limit our ability to conduct the experiment
by expanding it too much. Don’t want to prejudice opportunities in the
22. Abril i Abril:
Don’t agree with Kraaijenbrink that we want one or two of each category.
Don’t want to force the selection to satisfy diversity requirements.
More important to select robust proposals for TLDs.
23. Wilson: Calendar
must suit the numbers we are hoping for. Shouldn’t mislead ourselves
or the public. The current calendar can’t handle six to ten TLDs. The
current timetable constrains us to a smaller number.
• Roberts: Staff
and Board will have to discuss this on our monthly calls.
Have added language to proposal to reflect goals of diversity. Diversity
may not be most important point.
25. Conrades: Shoot
for quality not quantity. Make sure we can fulfill our promises to the
26. Murai: Important
efforts. Need to move forward.
27. Triana: Agree
on quality over quantity. Emphasize “measured and responsible.” Shouldn’t
try to exceed our means.
28. Wong: Concern
that schedule is too aggressive. But have great faith in the marketplace.
G. Voting on resolution as amended
1. All in favor,
IV. At-Large Elections
presented - McLaughlin
1. Concern expressed
that there will be only five At Large Members from 2001 to 2002. Have
a solution to clarify that the Board expects that there will be nine
At Large Directors at all times.
2. Also, want to
clarify that Study will consider all aspects of membership.
3. Wording is simple,
B. Wording - Sims
1. Suggest: “Remaining
four members of Initial Board shall serve until end of Annual Meeting
of the Corporation of 2002.” (Rather than 2001, as previously contemplated.)
• If any of the
four resigns and is replaced, the extension applies to their replacements
2. Re study, specification
of issues to be considered (others also, perhaps):
• Whether Board
should include At Large Directors
• How many
• How to select
them (At Large Membership, appointment by Board, selection by some other
• Proper role of
At Large Directors, processes and procedures for selection by Membership
• Proper structure
and role of Membership
3. Roberts: Note
that Board will charter the Study in LA. Plenty of time to discuss in
more detail later.
Wilson: Definition of vacancy?
• Touton: Yes, and
only for the duration of the original term.
5. Abril i Abril:
Are we increasing or decreasing confusion? And we’re extending terms
of the Initial Directors against previous promises.
• Wilson: It may
be the only alternative.
• Crew: Must take
notice of public comment. We’re adjusting our approach and Bylaws according
to public comments. This is what we should do.
• McLaughlin: If
the Board can act quickly enough to solve the problem after the study,
we could solve the problem faster.
• Wong: Don’t focus
on fact of “initial directors” but on skill set. There’s an upside to
keeping Directors around – creates an institutional memory that is helpful.
• Conrades: Crew
and I worked on Membership since the beginning. A hard problem. Election
and study will be informative. It’s good that we’re reassuring everyone
that we intend to learn from election process and do the study.
• Abramatic: Tradeoff
between building and changing. Need to be able to make changes while
sticking with fundamental commitments.
• Fockler: Have
all been thinking about this a lot.
6. Cerf: Move that
we adopt the resolution.
7. Dyson: Election
of first five Directors is a proof of concept. Happy to wait for the
last four. Changing your mind is no disgrace!
8. Triana: We’re
effectively asking what the role of one kind of Director should be.
That suggests there are two classes of Directors with different roles
and responsibilities. But if you rephrase the question to “proper role
of the Board,” then the study changes significantly.
• McLaughlin: Delete
the “proper role of such Directors” paragraph?
• Wong: But want
to save “taking into account of the limited technical and administrative
responsibilities of ICANN.”
• Sims: Move it
to preface of study so it applies to entire clause.
C. Voting on amendment to Bylaws:
All in favor, no abstentions
1. Abril: Vote in
favor of the package, but oppose one particular portion.
• McLaughlin: Minutes
will reflect that.
D. Election Process
1. Document presented
2. First resolution
is general approval of dates recommended by ElCom and staff. Second
resolution re implementation of rules. Third resolution re member nomination.
• Cerf: Need precision
in references – a title, not just a URL.
• Abril i Abril:
Ask about affiliations with national governments or international organizations.
Provisions in Bylaws prevent such individuals from being Board Members.
Bylaws already address this issue.
• Wilson: Request
a brief statement (not to exceed 250 words) re qualifications specifically
relevant to ICANN’s technical and administrative responsibilities, his/her
leadership and policy-level roles.
• Include in public
posting information about investments in Internet-related organizations?
Maybe too invasive?
• Dyson: Previously,
thought our conflict of interest statements should be public.
• Crew: ElCom recommended
not to post private phone numbers and email without the candidates’
• Cerf: Not sure
how I would have responded to that demand. List of consultancies, investments,
etc. is pretty invasive. The US Government doesn’t give that to the
public (though it does go to an Ethics committee). Should think hard
• Cohen: People
seeking to run for office should have to give “enough” information.
Have to figure out exactly what to disclose.
• Wilson: Every
board that I’m on has not demanded this in advance of serving. Rather,
conflict of interest policy binds once on the Board. Also, could have
a threshold – only disclose large interests.
• McLaughlin: Question
is what the community needs to know about a candidate to decide whether
the person has a set of affiliates that are troubling.
• Dyson: Shouldn’t
call it a conflict of interest. No one is saying that these interests
disqualify anyone. Policy is about disclosure, not disqualification.
Lost. Conflict of interest policy is clear, and we should ask potential
candidates whether they can meet the policy as it stands. But focus
• McLaughlin: No
conflict of interest to having, say, a registrar owner/operator serving
on the ICANN Board. Suggest inserting “ICANN-Related” throughout. Add
• Cerf: On track.
• Abril i Abril:
Too vague. What does “significant” mean?
• Roberts: Give
staff some direction. Don’t get caught up in details.
• Dyson: Use SEC’s
definition of “material” to specify the threshold.
• Crew: Board’s
intent is expressed in ElCom’s recommendation.
• Roberts: Vote
on Crew’s amendment?
E. ElCom Recommendations
1. How many candidates
to support? Geographic regions? Threshold of support required?
2. Fockler: Endorsements
limited to a Member’s region, but should be able to endorse multiple
3. Cerf: Should
be able to endorse multiple candidates. May not have decided yet which
candidate to vote for.
4. Roberts: Endorse
multiple candidates when can only vote for one?
5. Wilson: Don’t
want to have an upper limit on how many candidates a member can endorse.
6. Crew: If multiple
endorsements possible, may get too many candidates with the minimal
support level. Would then have to use other constraints to reduce size
7. Abril i Abril:
If limited to a single endorsement, giving endorsements requires thought.
If can make multiple endorsements, then the process becomes a game.
F. Threshold for support. Support
from x% of members of the region. Support by residents of at least two
1. Crew: Some favored
using a high level of support. Support from at least two countries has
not been disputed. Limited resources make it hard to handle many candidates
– likely max of 30 to 40. Recommend max of seven candidates per region,
combined between NomCom and Member-Endorsed candidates.
2. Roberts: So,
propose 2% of At Large Members in geographic region or 20 members (whichever
is greater). And support from members in two countries. Plus absolute
limit of seven candidates per region.
3. Wilson: Disparity
among regions in how many candidates to choose from.
4. Roberts: Note
that this isn’t intended to “squeeze out” member-nominated candidates.
5. Fockler: Disappointed
to have to cap the ballot. Endorsement phase becomes almost an election.
Unfortunate but I accept it.
6. Wilson: Should
be explicit that if we have too many candidates meeting the qualification
level, deciding factor for who goes on ballot is how many people endorsed
7. Cerf: ISOC has
a similar member-nominating process. Haven’t found it overwhelming.
But everyone who meets baseline criteria needs to get on the ballot.
8. Roberts: Hung
up on limit of candidates on ballot. And some are bothered by introducing
a primary in Member-Nominations.
9. McLaughlin: Less
sure that we’ll have overwhelming number of Member Nominations. But
don’t want to overwhelm the staff.
10. Wong: Extra
funds might help?
11. Roberts: It’s
hard to transfer skills to new human (or machine) resources. Doing so
would add a month (or more). Not pleased to be in that situation, but
think it’s reality.
12. Dyson: Think
the seven-candidate cap makes sense, not just because of staff resources
but voter source. Hoping to post extensive information about candidates.
To the extent that staff resources are stretched, should be for size
of Membership, not number of candidates.
i Abril: In only two months, Members must learn about ICANN, election
process, etc. Too much to ask for if there are thirty candidates. Even
seven seems like a lot.
14. Cerf: Want to
be clear about why cap is worrisome. If there is a cap, have to have
well-defined criteria for who makes it in. If number of endorsement
turns out to be huge, how to deal with the problem? Rules need to be
unambiguous and fair.
If number of candidates crossing threshold is less than number of available
spots, then they all get to be on ballot. If more, then take the top
n (when ordered by the number of endorsements).
16. Cerf: Process
is undefined if there are too many people making the threshold? (If
there’s a tie?)
Do it randomly. Or some principle. (Who got the nth endorsement first?)
18. Dyson: Case
seems unlikely. But if there’s a tie, we can let them all on the ballot.
19. Wilson: If there’s
a tie, pass the problem to the NomCom which could make a decision.
20. Cohen: Use “a
cap plus ties.” Unlikely that many candidates will have exact same number
of endorsements. Let’s move on.
21. Crew: In MAC
recommendation, suggested handling ties with random decision.
Agree with a cap. Favor “one who reached the tie first” approach.
23. Wilson: Stretching
staff resources. Maybe should drop back to 30 in case of ties.
24. Roberts: Staff
will have to support the election. If we don’t properly support the
candidacies, we will have a failure of trust.
25. Dyson: If using
exact numbers, possibility of a tie is very low. Ties more likely if
we use a percent threshold.
26. Roberts: Endorse
a candidate in another region?
G. Crew: ElCom recommended that
each member can endorse a single candidate in his own region.
1. Cerf: Attraction
to showing that a candidate has support beyond the region. Potential
danger is that some regions could influence the nominations in other
regions in a way that might seem unfair. On balance, oppose nominations
outside the region.
2. McLaughlin: It’s
been suggested that the next four At Large Directors be elected globally.
Doing that might address these concerns.
. Roberts: Place
a limit on number of member-nominated petitions? Clear consensus. For
unlimited? None. What should limit be?
• Dyson: The number
of Member-Nominated candidates on the region’s ballot. Five in a region
with five Member-Nominated candidates.
• Abramatic: Stick
with only one.
• Abril i Abril:
Allowing multiple endorsements facilitates some kinds of fraud. Safer
to limit it to fewer.
• Cerf: Abril i
Abril is persuasive. Need to determine whether software can implement
what Dyson wants.
Simplest solution is best. One. Let’s see how Board feels about that?
• Fockler: Agree.
• Wong: One. Makes
it easier to trace mistakes.
• Dyson: Three.
Provides some discretion.
• Wilson: Have to
make this simpler enough for people to understand.
• Roberts: Show
of hands for one? Most hands raised.
H. ElCom recommendations – two
V. Amendment to Bylaws
A. Roberts: Posted for extended
period. Little public comment.
B. Vote: All in favor. No abstentions.
VI. Subleasing of Additional Office Space
B. Cerf: Need more room. Prices
C. All in favor.
A. KPMG, recommended by Audit
Committee. Fee not to exceed $20,000.
B. All in favor.
VIII. Root Servers
A. Root Server System Advisory
Committee has come up with plans for a more secure way of updating the
root zone. Resolution
before the Board is authorization to move forward with it.
B. Cerf: Board should review proposal
1. Roberts: Difficult
to convene the Board in August.
2. Cerf: “Shall
be made available to the Board prior to submission to the DoC…”?
IX. Geographic Regions
A. See http://www.un.org/Depts/unsd/methods/m49regin.htm
B. Cerf: Make sure the list is
cited in a way that persists.
C. Abril i Abril: Concerned about
telling countries what regions they’re in. There’s no such thing as
a neutral list.
D. Dyson: Have to make the system
work. Are pleased to have delegated responsibility to someone else.
E. Cerf: Have to make a choice.
Reconsider during study.
F. Triana: Congratulations to
staff for finding a solution to this complicated problem. It’s good
that this list fits so neatly with bylaws.
G. Touton: List includes codes
that are not countries (areas).
1. McLaughlin: Areas
should be grouped with countries. A person who lives in Guadalupe would
be classified with France.
H. Duration of policy?
1. Touton: Bylaws
provide that Board will review geographic assignment at least every
2. Roberts: Can
limit policy to the election. But need to make a decision now at least
for the election.
I. Cerf: Consistency with RIRs’ guidelines.
J. Abril i Abril against. Wong
abstaining. Rest in favor.
X. Voting System Update
A. Resolution authorizes President
working with Staff and ElCom to select a voting system vendor. Rapid
B. Triana: Notification to the
1. Roberts: Sure.
C. Roberts: Note tradeoffs between
financial terms, system specifications, etc. System has to be able to
handle 60,000 votes. Some vendors have already supported larger elections
D. Voting: All in favor.
XI. Status of members.icann.org server
A. “Load management”
B. How we got to this “awkward”
1. Discussion in
Santiago re how to move membership forward. Proposal submitted to Markle
for $100,000. Took longer to obtain than we expected. Announcement of
grant made only in November.
2. Board’s decision
in LA contemplated an indirect method. Expected 1,000 to 10,000 voters.
Wanted a system that could support 100 applications per day, peak of
500 per day. Those were the loads received in February and March.
3. Application rate
increased from ~500 per day to consistently more than 1,000 per day,
as much as 2,000 per day. Created a huge membership base (already 42,000
verified, more than 50,000 unverified). Expect 90,000 or more registrants.
4. Concerns about
“fairness” of server not being responsive at all times of the day. Staff
has no opinion.
5. Thanks to Crispin
who helped with improving system.
6. After July 31,
system use will be less database-intensive. Improvements in August may
C. Dyson: Now we’re forewarned
re planning for capacity. But what we can we do in the next two weeks?
Mini-application to take their names and email addresses for subsequent
follow-up? People who want to register should be able to. Maybe a separate
1. Roberts: Network
access is not a problem. Issue is interaction of dynamically-generated
web pages with database. Outside experts say little can be done. Every
record in the database creates requirements for letter-printing, etc.
that have to be satisfied, so can’t readily move to a new database.
Have decided that it’s better to completely serve the people who have
already registered rather than to do a bad job for everyone.
2. Cerf: Additions
to what’s already there risks fouling up the entire process. Nervous
about making changes at this point. Have to deal with this “success
3. Dyson: Suggest
a separate application to receive registration requests.
4. Wilson: We’re
all grateful for outreach. But have to be able to assure that the system
5. Sims: The system
must be working well if it registered ten times the people it was expected
6. Abril i Abril:
Add mailing list to members page for those who can’t get registered
to be full members?
• Roberts: Outreach
task force hasn’t made a recommendation yet re how to proceed once Members
site stops receiving new applications. Possibility of referring users
to the Internet Society.
7. Murai: Agree with Cerf
re scaling and risk. Not changing anything for the same reasons Cerf
notes. Also, Japanese people suffer from the invalid translations to
8. Roberts: Also,
note that hosting the Root will entail substantial costs, likely on
the order of $250,000 per year.
9. Roberts: System
is performing quite well on the basis of the specifications we had agreed
XII. At Large Director Selection Plan
A. Review of ElCom recommendations
B. Cerf: Concerned that voter
education and candidate communication processes limit ability of some
candidates to conduct a campaign.
1. Roberts: Before
candidates get to this point, they have to have activated their membership.
Anyone who can be a candidate already has web access.
2. Dyson: But what
it it’s hard for a candidate to get web access?
C. Cerf: Concerned that this process
will turn ICANN into something it isn’t.
1. Roberts: Well-meaning
people all over the world are mistakenly calling this a “global election”
and not noticing that we’re a small company with a limited mission.
D. For Monitoring and Oversight,
choice between Independent Individuals versus Election Observation Institutions.
E. Fockler: In large regions where
the 2% is greater than 200, a candidate only needs 200 supporters to
meet the threshold. Threshold seems non-threatening.
F. Cerf: This bugs me too. Huge
disparities among countries within a region are troublesome. But how
to deal with this problem?
G. Fockler: Uncomfortable with
telling someone with huge support (~500 people) that he or she can’t
be on the ballot. Used to being able to get on the ballot; uncomfortable
with being pushed off the ballot based on an effective primary election.
H. Abril i Abril: Not holding
an election, but trying to select a highly-qualified slate of board
I. Wilson: To accomplish Fockler’s goals, would
need to adopt additional protections based on nationality also. Not
inclined to do that.
J. Kraaijenbrink: Thought we were
looking for candidates with most support. Doesn’t matter if there’s
a bottom threshold or not. Stick with old proposal.
K. Cerf: Equal access to the ballot
important. Stick with old proposal.
L. Murai: Don’t focus on nationality.
M. Mike: little support for Fockler
N. Fockler: not nationality focused
O. Voting on resolution as amended:
Fockler abstains, others in favor.
XIII. Thank You Resolutions
1. Adopted unanimously
For additional technical information, please contact:
Ben Edelman and John
Berkman Center for Internet & Society at Harvard Law School