IUF 2006 03 26

Jump to: navigation, search

[edit] Synopsis

We'll distill the stream of consciousness notes from the sections below into something short and pithy that will be useful for the folks who can't be here in person.

[edit] Agenda

Image:Brett-tld.JPG

Bret Fausett moderates.

  • Welcome and Introductions
  • What are the lessons learned from previous rounds of domain name selections and how do we apply them?
  • How do we "internationalize" top level domains?
  • What sould the next Top Level Domain application process look like?

Image:Elliot-tld.jpg

Elliot Noss gesticulates on the subject of market forces.

Image:nz-tld.jpg

The NZ registry team staunchly advocates the value of community input.

[edit] Participants

[edit] Notes by Raymond King on new TLDs

  • Dirk Krischenowski - There should be new TLD's and believes all proposals should be considered as potentially possible
    • example, .DE has almost 10MM already and the space is getting tight.
    • ALAC has done good work in showing why there should be more TLDs
  • John Berryhill - The perception in the intellectual property community is that new TLDs provide opportunities for Cybersquatting / Every new TLD is a TM shakedown
  • Bret Fausett - some view every new TLD as a way for new registries to hit them up for protection money and get nothing of value
  • John Berryhill - his son thinks .XXX is a good idea, but wonders why unrelated trademarks such as Pontiac.xxx would be an issue - why would anyone look there for automotive information? The idea that trademark owners have to buy their names in new TLDs is somewhat of an issue of how its been sold, but who is going to be looking for CocaCola.whatever instead of CocaCola.com?
  • Brett - that may be true for .orchid, but what about something more generic, like .web?
  • John - Trademark law is not a pre-emptive mechanism, but a remedial system / look at gaming in .eu, Benelux Trademarks is an example of people using the trademark system just to get names. .biz stop procedure worked well in comparison to sunrise methods.
  • Dirk Krischenowski - .Berlin is focused on the city community of Berlin. Its an opportunity to create a short, clear an concise identifier. They will have sunrise period.
  • Elliot Noss .biz solution worked for TM owners on a limited basis. Sunrise in .biz and .info crippled those TLDs. There is a way to protect TM without crippling the Namespace. Insiders gaming sunrise and allocation process and selling them later does nothing for the Namespace.
  • Elliot Noss Relics that need to be eliminated
    • Fixed price
    • First Come, First served
    • scarcity of TLDs
      • anything we do that is a half measure will lead to the same problems.
  • Bret Fausett how do we get the names into the hands of people who are going to use them and not in a TM-infringing way
    • one idea that's come up is auctioning. if someone is going to pay $500K for a name, then they are likely to use it and they are not likely to pay a lot and lose a name in a TM dispute.
  • Paul Stahura autioning good, but who gets the money?
  • Elliot Noss - auction the TLD, then the registry who wins that auction can auction the second level names.
  • Someone from audience Rodney Prescott - InternetNZ- auction process would put other countries at a disadvantage vs. western countries, and/or more wealthy countries or even companies. It changes the game.
  • Panelist - so long as domains are transferrable at any level, there will be an auction. It just boils down to who gets the money. Beauty contest has had poor results to date. Suggestion - put enough names out there that there isn't as much contention.
  • Dirk Krischenowski necessary to have auction in order for small TLD to pay investors back for startup costs.
  • Elliot Noss - we're all stuck with what we know / browsers, e-mail. expanding what's to the right of the dot adds semantics to the right of the dot to return the namespace to what its supposed to be, which is an aid to navigation.
  • Bret Fausett - There is no one size fits all for TM protection in TLDs
  • Elliot Noss If we image IBM publishing all of it's information at .IBM names, then we don't have to worry about people squatting at ibm.whatever.
  • Panelist - problem with descriptive names that create directories, like .museum, is that the search engines are so good now, that we don't need the "directory TLD". .aero is another good example, not that many have been reserved so far. With .aero, almost all of the names are just re-directs. .Coop is another good example to look at, where the names mostly redirect.
    • .edu works like an STLD
  • John Berryill On why partial solutions are good: "Condoms aren't 100% effective, but what's the alternative, tell people not to have sex?"
  • Paul Stahura on why some of the TLDs didn't work out well
    • you had to exaggerate your projected numbers
    • in order to win the TLD, you had to hobble it
    • so, I'm (paul) for the auction, maybe there is a tld that works with just 200 names in the zone
  • Audience - once you make it easy to make new TLDs, what is the potential market? Is there a technical problem to have 10,000 TLDs?
  • Bret Fausett - I don't think the stability is an issue with the kind of numbers of new TLDs we're talking about.
  • Elliot Noss paper at DNS Policy Paper
  • Wendy Seltzer Let the market decide -- participants can vote with their dollars whether they think new gTLDs are useful
  • Elliot Noss qualifying some said before. In ccTLDs, there is in fact lots of meaning to right of the dot.
  • Dirk - Semantic TLDs are in fact likely to rank websites better at search engines, we had talks to Google and Yahoo about this
  • Wendy Seltzer Innovators dillemma, we don't know what will be successful, .cantelope may be successful, who knows?

[edit] Notes on IDN by Raymond King

  • use dName and let the incumbant registry take care IDN
  • use NS records
  • from what Paul Twomey said yesterday, what is the definition of a ccTLD?
  • there will be applications for .IDN something
  • This is a minefield and i'm not certain that an auction system is going to solve the problem because the American winning bidder for a chinese TLD will be a political issue, etc.
  • Audience - .com is ambiguous to some, is it "COMpany" or "COMmercial" ?
  • Dirk - dNames for .berlin may not have as much meaning in other countries, but we think of using this option
  • John Berryhill IDNs don't add a new dimension to TM issues. TM's are about consumer perception, not the actual name.
  • Bret Fausett thinks there will be tremendous demand for IDN TLDs
  • John Berryhill .iii had the benefit that it didn't mean anything and therefore couldn't be offensive.
  • GNSO yesterday (as reported by Bret Fausett)
    • new applicants should have technical ability (but no agreement on what it should be)
    • have minimum amount of $ behind you (but no agreement on how much it should be and how to measure it)
    • should you need a business plan - then looks like a beauty contest (no real idea as to how to judge it)
    • people want to move to a more objective process and get ICANN out of the beauty contest business.
  • Paul Stahura if it is an auction, the VCs really do decide since they will put $ behind many of the applications - maybe we should let the VCs decide!
  • Wendy Seltzer Markets have proven over time to be a better way to do allocation, vs. gov't processes. Especially when we don't really know what users want, etc. Money serves as a proxy for what people believe is worthwhile
  • Elliot Noss Wendy talks for the public good (not an evil business perspective!). The market will decide better than ICANN or a council of elders.
  • Audience - I don't agree with that, it will develop along the lines of what makes the most money for some people
  • Audience (first time ICANN guy -- (that would be me, Mike O'Connor)) - I represent an experiment, I own corp.com domain. he's planning to launch between now and Morrocco ICANN (well, fingers crossed... M). I like the market deciding, let's see what happens.
  • Tim Ruiz - there are going to be failures, that's just a fact of life, let's put some mechanisms in place to recover from those failures. It's not realistic to expect that there won't be any failures.
  • Audience - how can we say there's no need for something, when others may think there is a need. To say from the top "there's no need" may be too strong.
  • Bret Fausett - we've talked about some TLDs that shouldn't go to the highest bidder, but rather the most appropriate bidder. And likewise some TLDs that should be allowed at all.
  • Annette Muehlberg some domains may not be appropriate, but if someone has $ to buy it that might be a problem in the purely market driven approach

[edit] Notes by BrandonCsSanders

  • Welcome and Introductions
    • We'll be talking about gTLDs (generic Top-level domains like .com, .net, .org, .info, ...)
    • The GNSO is also talking about this same issue in parallel
    • Question before the house ... what should the new gTLD process look like?
      • What should the call for proposals for gTLDs look like?
      • How should they be approved ...
      • What did I miss here? Ack
    • Famous and really smart folks on the panel are introduced
    • Should we have new gTLDs?
      • Yes ... consensus
      • But what yes looks like varies dramatically across gnso constituencies
      • e.g., no new broad ones like .web
      • or for e.g., only sponsored ones like .jobs
    • Dirk says yes ... with no limitations for te kinds of tlds ... no reason to exclude applications ... approval might be more limited. Should be open. Really hard for new companies to get a short name that is representative.
    • Brett puts john on the spot "put on your IP hat and say why there shouldn't be more new tlds"
    • John squirms and says "Consider the IP constituency doesn't want new tlds because it provides new opportunities for cybersquatting. The fear is that Ford.orchid would confuse people. Every new tld is an opportunity for registry operator to shake down poor multi-nationals in a protection racket"
    • Brett ... companies got nothing of value from new tlds just because they have to register domains they aren't going to use
    • John ... son and unrelated comments about .xxx and folks will register pontiac.xxx, the perception that trademark owners are being shaken down because registrars/registries tell them that they have to. If you are looking for coke you go to coke.com rather tan coke.xxx or coke.orchid so registering them doesn't add any value for coke.
    • One argument that we've heard from bigguns is that they are less worried ab out specific ones like .clothes or .jobs because they don't worry about the mistakens .web and general ones like that are the bone of contention
    • Many comments
    • John speaks about trademark law as a remedial versus a proactive ... use the trademark system as a way to get domains they want ... 60% of the .biz stop claims were decided in favor of those who registered
    • John thinks the .biz stop bizness worked better than the .info sunrise
    • Dirk and .berlin will generic trademarks will try to apply during the sunrise period
    • Eliott put on the spot ... what worked in .biz the sunrise was the single biggest crippler of the info and .biz spaces by shelving the most desirable names from the outset
    • John thinks Eliot is mixing things but we'll get back to that
    • Brett .info had the sunrise, .biz had the stop
    • Intense exchange ... about whether the sunrise/stop
    • Allocation as the problem?
    • IP constituencies say they hate to buy the domain names they are never going to use, the registries and registrars don't like them being shelved, and the users don't get to actually register them
    • Elliot speaks of religious artifacts, "scarce namespace", flat pricing model
    • Dirk: From marketing side, .biz and .info had little marketing budget and marketing (asserts need 100 million, only have $500,000k or a $1 million) ... need reference domains that would make it more visible
    • Brett ... one solution is to get names in the hands of those who will actually use them in a non-infringing matter ... one way would be to have second level domain auctions
    • Anyone an advocate for te auctions
    • Paul thinks its a good idea ... just to reinforce ... if you are going to auction them folks aren't going to buy trademarked names in the auction
    • So who gets the money?
      • If I had a tld it would be me
      • The money goes to the registry
    • You auction the tld, then the registry gets to auction the second level domains
    • Some other nations are going to be significantly disadvantaged by the auction process
    • How to represent te interests of emerging economies?
    • Where does the money go?
      • Take the money and subsidize the developing countries
    • John, so long as domains are transferrable there will be an auction
    • Either you have an auction, or you have a beauty contest
    • ICANN is a bad destination for the lucre because there isn't good oversight
    • speculators aren't good because you don't get economic value
    • beauty contests haven't worked well
      • I can have a poor virtuous company front for me
    • Game playing with the beauty contest
    • Lawyers win
    • Prefer the plan of hyper inflation so that te value goes away
    • Dirk, the auction at .berlin had to front 2 million euros ... absolutely necessary to hold an auction to pay back the donors
    • Folks remember the usenet arguments about starting newsgroups
    • Any body could create a
    • Could be that the information gets hard to find
    • There currently isn't any semantic content to the righ of te dot
    • We're not going to nail this coming out of the gate
    • There are lots of folks on the inside who know what to do with the money
    • Confident there will be many more good ideas than bad ideas
    • http://dnspolicy.org
    • Let the tld's choose what mechanism they use (sunrise or stop or ...)
    • IP and business constituencies have asked that every new tld have mechanism to deal with the trademark issue
    • Having every new tld have a different IP policy is a drag, 50 new tlds, each one having a different policy
    • There is no one size fits all
    • If we imagine ibm publishing all of their information in .ibm
    • stlds, looked at the number of registrations, they all got less than 1/20th of their most pessimistic estimates
    • .museum and categorizing via top level categories, it just didn't work ... they did a great job and "who cares", does anybody type in a domain name? no, they search
    • .aero they reserved all the two letter airline codes and all the three letter airport codes ... very few have actually been claimed
    • .aero is about as natural of a natural directory as you could have ... the industry doesn't care
    • .coop ithaca is awash in coops, it doesn't matter because being a coop is not a big deal ... coop did everything right, it's not meaningful branding
    • .museum, .aero, .coop had very little uptake even within their target communities
    • We've used lots of ICANN staff time on these smaller issues ... couldn't we better prioritize by tlds that will serve millions?
    • Disagree ... .edu works, because the difference between a real educational institution versus a fake one isn't very large
    • What would .bank get out of it?
    • Partial solutions are good
    • Paul, in beauty contest incentive to exagerate, and perhaps they didn't work out because of all the hobblings they had to do

[edit] Notes on the Auction Process by BrandonCsSanders

  • If there are 10,000 new tlds there isn't a
  • Is there really a demand for the tlds?
  • Just creating something doesn't mean people will come
  • The right of the dot is not going to have meaning ...
  • Wendy: Why are we thinking about about what is going to be best for the end users ... lets let the end users decide for themselves
  • How did the nz folks decide they didn't want to have .bank.nz
    • We had a vote ... publicized, discussion on mailing lists, isoc etc
  • What selection procedure scales?
    • Just publishing on a public website isn't enough
  • AUDA lady in red shirt ... had twelve different possibilities, none of them got legs
    • How do you determine whether they have legs?
    • Due process and people follow it ... clear process
    • Really interesting arguments
  • What if we had the process (bring me a zone file with 100,000 registrations) then we can talk
  • Have we got a level of interest that justifies the process?
  • Localized search and the search engines ... semantic tlds will help you get better rankings in search engines
  • Wendy returns to "Let the market serve as the decision process" ... innovators dilemma, maybe 99 out of 100 don't work, but those that are successful might be successful for reasons we can't even anticipate
  • Concerns for the registrants who register
  • Transition to IDNs

[edit] IDNs by BrandonCsSanders

Protocol floated by the registries

  • Afilias has .info, rather than have them rebid, info.pr would automatically get it
  • Another version is have .com.pr open for bid
  • What is a cctld?
    • Twomey, the cctld is only the entries in the iso list
    • When we launch the new .idn process later this year who gets to win?
    • Winning bidder for a .pr tld coming from the us would be a politically hot potato
  • What does the idn roll out mean to you?
  • .nz
    • Maori language is a problem
    • Doesn't seem like


ICANNwiki: An industry resource fostering global collaboration and transparency within the ICANN community

Platinum Sponsors

Skenzo
LogicBoxes
Job Board

Gold Sponsors

silver Sponsors
Bronze Sponsors
Past Conferences
Paris
New Delhi
Los Angeles
San Juan
Lisboa