ICANN lists the ways the new gTLD program sucked

ICANN has published an analysis of the many ways in which the first round of the new gTLD program wasted everyone’s time and money.
The 200-page “New gTLD Program Implementation Review” is essentially a long list of ways the program could have been better, along with dozens of recommendations for possible future changes.
It’s for the most part a fairly dry read, and it is probably not as comprehensive as it could be, but it will be required reading for anyone working on policy concerning, or thinking of applying during, the second application round.
It concludes, for example, that maybe there should be a right to appeal inconsistent objection rulings.
It ponders aloud whether the Community Priority Evaluation should be scrapped or revised.
It wonders whether dot-brands, or other categories of gTLD, should get their own version of the standard Registry Agreement.
There’s also some discussion about the possibility of making the evaluation stage more efficient by grouping applications by applicant or back-end service provider, which would streamline the process but complicate the prioritization queues.
I count 48 “lessons learned” in the document, but as a concise summary covering over three years of the program, it’s necessarily somewhat light on detail.
On my first read, a few omissions jumped out at me.
There’s no discussion at all of the cybersquatting component of the background screening process, for example. Nor is there any mention of Geographic Name Review shortcomings highlighted by the recent .africa Independent Review Process case.
Also, in my view the document goes way too easy on the Governmental Advisory Committee.
That’s just off the top of my head. I’m sure almost everyone who reads it will notice something lacking.
That’s why it’s now open for public comment.
The document is expected to be used as part of the review leading into the second application round, which somehow seems more distant with each passing day.

US gov: we can’t support ICANN accountability plan

The US National Telecommunications and Information Administration has waded into the ICANN accountability debate, possibly muddying the waters in the process.
In a blog post last night, NTIA head Larry Strickling said that community proposals for enhancing accountability were not yet detailed enough, and had not reached the desired level of consensus, for the NTIA to support them.
He urged everyone involved to simplify the proposals and to work on areas where there is still confusion or disagreement.
The comments were directed at the Cross Community Working Group on Enhancing Accountability (CCWG), a diverse volunteer committee that has been tasked with coming up with ways to improve ICANN accountability after the US government severs formal oversight of the IANA functions.
That group spent a year coming up with a set of draft proposals, outlining measures such as stronger, harder-to-change bylaws and improvements to the Independent Review Process.
But the main organizational change it proposed is where the most conflict has emerged.
CCWG thinks the best way to give the community a way to enforce accountability is to change ICANN into a membership organization, a certain type of legal entity under California law.
It would have a Sole Member, a legal entity peopled by members of each part of the community, which would have to right to take ICANN to court to enforce its bylaws.
The ICANN board doesn’t dig this idea one bit. Its outside attorneys at Jones Day have counseled against such a move as untested, overly complex and potentially subject to capture.
On a recent three-hour teleconference, the board proposed the Sole Member model be replaced by a “Multistakeholder Enforcement Mechanism”.
The MEM would create a binding arbitration process — enforceable in California court — through which ICANN’s supporting organizations and advisory committees could gang up to challenge decisions that they believe go against ICANN’s Fundamental Bylaws.
Since this bombshell, a key question facing the CCWG has been: is the board’s view being informed primarily by its lawyers, or has Strickling been quietly raising NTIA concerns about the proposal via back-channels?
If it’s the former, the CCWG and its own outside counsel could robustly argue the community’s corner.
If it’s the latter, it’s pretty much back to the drawing board — because if the NTIA doesn’t like the plan, it won’t be approved.
Unfortunately, Strickling’s latest blog post avoids giving any straight answers, saying “it is not our role to substitute our judgment for that of the community”.
But his choice of language may suggest a degree of support for the board’s position.

As I stated in Argentina in June, provide us a plan that is as simple as possible but still meets our conditions and the community’s needs. Every day you take now to simplify the plan, resolve questions, and provide details will shorten the length of time it will take to implement the plan and increase the likelihood that the plan will preserve the security and stability of the Internet. Putting in the extra effort now to develop the best possible consensus plan should enhance the likelihood that the transition will be completed on a timely schedule.

The emphasis on “simplicity” could be read as coded support for the board, which has repeatedly said that it thinks the Sole Member model may be too complicated for the NTIA to swallow.
Both the board and Strickling’s latest post refer back to a speech he made in Buenos Aires in June, in which he said:

If a plan is too complex, it increases the likelihood there will be issues that emerge later. Unnecessary complexity increases the possibility that the community will be unable to identify and mitigate all the consequences of the plan. And a complex plan almost certainly will take longer to implement.

Strickling certainly knows that the board has been citing these comments in its objection to the Sole Member model, so the fact that he chose to repeat them may be indicative of which way he is leaning. Or maybe it isn’t.
Either way, I think it’s going to be tough for the CCWG to easily dismiss the board’s concerns.
CCWG members are currently on planes heading to ICANN headquarters in Los Angeles for a two-day face-to-face meeting at which the chairs “expect that a large portion of our time… will be reserved to answering the tough questions”.
Many believe that unless this meeting is extraordinarily successful, it’s going to be tough for an IANA transition proposal to be approved by the NTIA under the current US administration.

XYZ to rethink China gateway plans

XYZ.com has withdrawn its request to start selling .xyz and .college domains into China via a local gateway service provider.
The company has said it will amend and resubmit its plan to ICANN, which had told it the idea “might raise significant Stability or Security issues”.
The registry wants to be one of the first non-Chinese registries to be able to comply with government regulations, which require all domain firms to have an official license.
As we reported last week, it had signed up with local registrar ZDNS, which would proxy for registrations made by Chinese registrants.
However, it has now withdrawn its Registry Services Evaluation Process request after ICANN said it would have to refer it up the chain to a special technical committee for review.
XYZ said in a letter to ICANN:

We are withdrawing this request because our gateway model is changed since the submission of the registry request and so the request is no longer accurate. We will shortly submit a new registry request to cover the updated gateway model.

It’s not clear what the specific “security and stability” concerns were.

.sexy may be blocked in Iran

Some networks in Iran appear to be systematically blocking Uniregistry’s .sexy gTLD.
That’s one of the conclusions of a slightly odd experiment commissioned by ICANN.
The newly published An Analysis of New gTLD Universal Acceptance was conducted by APNIC Labs. The idea was to figure out whether there are any issues with new gTLDs on the internet’s DNS infrastructure.
It concluded that there is not — new gTLDs work just fine on the internet’s plumbing.
However, the survey — which comprised over 100 million DNS resolution attempts — showed “One country, Iran, shows some evidence of a piecemeal block of Web names within the .sexy gTLD.”
The sample size for Iranian attempts to access .sexy was just 30 attempts. In most cases, users were able to resolve the names with DNS, but HTTP responses appeared to be blocked.
The survey did not test .porn or .adult names, but it might be safe to assume similar behavior in those gTLDs.
APNIC also concluded that Israel’s .il ccTLD, included in the report as a known example of TLD blocking at the national level, is indeed blocked in Iran and Syria.
The study also found that there may be issues with Adobe’s Flash software, when used in Internet Explorer, when it comes to resolving internationalized domain names.
That conclusion seems to have been reached largely because the test’s methodology saw a Flash advertisement discretely fetching URLs in the background of web pages using Google Ads.
When the experimenters used HTML 5 to run their scripts instead, there was no problem resolving the names.
The study did not look at some of the perhaps more pressing UA issues, such as the ability for registrants and others to use new gTLD domain names in web applications.

Sunrise accounts for under 1% of new gTLD regs

New gTLD registries can expect just 125 sunrise registrations on average, according to statistics just released by ICANN.
The new data, current as of May 2015, also shows that there have been just 44,077 sunrise registrations in total, over 417 new gTLDs.
That’s less than 1% of the total number of new gTLD domain registrations to that date.
The numbers were published in a revised version of ICANN’s Revised Report on Rights Protections Mechanisms, a discussion paper on mechanisms such as sunrise, Trademark Claims and URS.
It also contains the first authoritative breakdown of sunrise regs by TLD, though it’s limited to the 20 largest.

Many of these numbers match closely what DI has previously reported, but .porn and .adult are substantially lower because ICM Registry only revealed consolidated numbers that took account of its unique non-TMCH sunrise periods.
None of the ICANN figures include .sucks, which hit sunrise after the numbers were compiled in May.

ICANN throws lifeline to flopping new gTLDs

New gTLD registries with lower than expected sales will now be able to reduce the amount of their “failure bond”.
ICANN has introduced a new Continued Operations Instrument Amendment Service, which will enable registries to raise or lower the amount of their COI depending on how business is going.
A COI is a letter of credit or cash in escrow that registries must secure in order to fund three years of emergency operations in the event that their businesses fail.
The amount of the COI is calculated from sales projection and ranges from $18,000 (for under 10,000 names) to $300,000 (over 250,000 names).
Let’s face it, at the moment the amendment service must surely be targeted largely at companies that over-estimated their future sales and secured a COI much larger than they needed.
If they’ve escrowed cash, the new service will allow some of that money to be freed up to spend on more useful activities.
ICANN said that if it determines that a registry has under-projected its sales, it will be able to refer it to the new service in order for the COI to be increased.
Currently, only four new gTLDs have over 250,000 names under management, judging by zone files.

ICANN just gave a company a new gTLD for free

The Tor Project Inc, a Massachusetts non-profit software maker, just got a new gTLD reserved for its own exclusive use, by ICANN, for free.
Tor did this without engaging in the ICANN new gTLD program, paying any ICANN application fees, or following any of the rules in the ICANN Applicant Guidebook.
It basically circumvented the entire ICANN process, and it only took six months from asking.
Neat trick, right?
Tor develops the software that creates the Tor “anonymity network” used by people who wish to obfuscate their internet usage (legal or otherwise) by routing their traffic via a series of proxies or relays.
The free software, which plugs into browsers, uses meaningless, hashed “.onion” domains because the routing method is known as “onion routing”.
IANA, an ICANN department, last night placed .onion on its list of Special Use Domains, meaning it cannot be delegated to the DNS.
If anyone were to apply for it today — assuming that were possible — they’d be out of luck. It seems .onion now has the same protected status as .example and .localhost.
The reservation was made at the instruction of the Internet Engineering Task Force, which published a new Internet Draft reserving the .onion gTLD for use with Tor.
An Internet Draft is a “work in progress” standards track document with a six-month shelf life, not yet a finalized Request For Comments (RFC).
This one was written by engineers from Tor and Facebook.
The Internet Engineering Steering Group, the IETF’s coordinating body, approved the draft last week.
Of the 13 IESG members who voted on the document, the first draft of which was published six months ago, five voted “Yes”, seven offered “No Objection” and only one abstained.
The abstainer, Barry Leiba, standards guru at Huawei Technologies, wrote:

I believe the IETF shouldn’t be involved with registering special-use TLDs for things that were used outside of IETF protocols, and should not be wading into territory that belongs to ICANN. I know there are a bunch of other such TLDs that people/organizations would have us snag for them, and I very much want to avoid doing a batch of others.
That said, I well understand the deployed code involved and the importance of keeping things working in this case, and I don’t want to stand in the way. So I’m standing aside with an “Abstain” ballot.

The logic behind the reservation is that if ICANN were to delegate .onion to somebody else (for example, The Onion) there would be a risk that the improved privacy offered by Tor would be compromised.
Voting in favor of the draft, Cisco engineer Alissa Cooper wrote:

Registering this name seems warranted in light of the potential security impact. We need to make our processes work for the Internet, not vice versa.

Another affirmative vote came from Oracle engineer Ben Campbell. He wrote:

This one took some soul searching. But I think the arguments have been made, and that on the whole this registration does more good than harm.

A number of IESG members suggested that the IETF should revisit and possibly amend the RFC in which it originally granted itself the power to reserve gTLDs.
That’s RFC6761, entitled “Special-Use Domain Names”, which dates to February 2013.
RFC6761 lays out a seven-point test that a string must pass before it can be considered “special use” and thereby reserved.
The tests cover whether humans, applications and various types of DNS software are expected to handle the string differently to a regular TLD.
The RFC also notes:

The IETF has responsibility for specifying how the DNS protocol works, and ICANN is responsible for allocating the names made possible by that DNS protocol… Reservation of a Special-Use Domain Name is not a mechanism for circumventing normal domain name registration processes.

I think reasonable people could disagree on whether that’s what has just happened in the case of .onion.
Indeed, there was some discussion on the IETF’s “dnsop” working group mailing list about whether Tor was “squatting” .onion, and whether it was appropriate to reserve its chosen TLD string.
I wonder what kind of precedent this could set.
The Tor Project Inc is a Massachusetts non-profit company. It’s primarily funded by US government grants, according to its 2013 financial statements, the most recent available. It doesn’t sell .onion domains — they’re auto-generated by the software.
Part of the argument in favor of allowing the new Internet Draft is that .onion substantially pre-dates the creation of RFC6761 — it’s not an attempt to game the RFC.
Why wouldn’t that same argument apply to, for example, alternate root operator Name.Space, which has been offering hundreds of pseudo-gTLDs since 1996?
Name.Space could argue that its strings pre-date .onion by eight years, and that the security of its registrants and users could be compromised if ICANN were to delegate them to the DNS.
What about NameCoin, another alternate root provider? It also pre-dates RFC6761 and, like Tor, uses browser software to work around the DNS.
I don’t know enough about the IETF’s processes, to be honest, to say whether it would be forced to apply its .onion logic to these other namespaces. But it’s an interesting question.
And as somebody who has spent the last five years immersed in the minutiae of the rules ICANN has created to govern the allocation of words, it’s jarring to see those rules circumnavigated so completely.

Who wants ICANN’s $60m gTLD windfall?

ICANN has opened a formal public comment period to move forward discussions on how it should spend the almost $60 million it has so far received in new gTLD auction proceeds.
It’s not yet looking for concrete suggestions on how to spend the money — this is a pre-consultation consultation — it’s only looking for comments on the principles that should be considered when discussions take place.
ICANN has so far raised $58.8 million from “last resort” new gTLD auctions. With 27 contention sets remaining, that number could go up if one or more applicants refuse to participate in private auctions.
The GNSO Council has been moving to create a Cross-Community Working Group to discuss how the money should be spent, but clashed briefly with the ICANN board, which has said it will make the ultimate decision, earlier this year.
The new paper (get it here) basically asks questions along the lines of: who should decide where the money goes? How should conflicts of interest be handled? How much third-party expert opinion should be solicited? How much say should the board have? How much outreach should there be?
Underpinning it all is the implicit problem that the longer, more detailed and more convoluted the process, the less money there will be to actually distribute at the end.
Knowing the ICANN community’s propensity for convolution, I wouldn’t be surprised if it managed to spunk the whole lot on expert advice, working group travel, lawsuits and coffee.
(Okay, I would actually be surprised, but you get my point).
The paper also includes links to about 20 spending suggestions that have been made in various public fora over the last couple of years.
Some ideas include: giving it back to the applicants, funding open source DNS software, reducing the new gTLD application fee, marketing new gTLDs to registrants, and donating it to charity.
It does not appear to be true that ICANN slipped in one of its own management’s suggestions in an attempt to funnel off new gTLD money into the unpopular NetMundial initiative, as has been alleged elsewhere today. The NetMundial suggestion referred to in the paper actually came from Danny Aerts of Swedish ccTLD manager IIS.

XYZ fighting red tape to serve Chinese customers

XYZ.com is trying to become one of the first non-Chinese gTLD registries to be able to sell unhindered into the Chinese market, in the face of Draconian government regulations.
The company has filed a Registry Services Evaluation Process request with ICANN — the first of its kind — that would let it use a gateway service, based in China, to comply with strict local laws on registries, registrars and registrants.
The Ministry of Industry and Information Technology regulations have been in place for a decade, but it’s only in the last year or so, in light of the new gTLD program, that China has been strictly enforcing them.
Anyone in China can buy a domain, but you need a license if you want to put up a web site, according to Gavin Brown, CTO of .xyz back-end CentralNic. Registrants also need to have their Whois information verified and validated, he said.
The problem for Chinese residents today is if they buy a domain in a TLD that is not licensed by the government, they won’t be able to obtain a license to host a web site on that domain.
The .xyz gTLD is believed to have a few hundred thousand domains registered via Chinese registrars, a substantial portion of its total.
There’s a worry that China could demand the deletion of these names and others, as it has previously in .cn, if the proper licenses have not been obtained.
Naturally, the inability to use these domains has led to a lot of pissed-off registrants. XYZ says has been receiving complaints from its registrars in the country, which in turn have been receiving complaints from their customers.
XYZ proposes to fix the problem by using a gateway service provided by ZDNS, a DNS provider based in mainland China.
Registrars in the country would maintain a separate EPP connection to ZDNS, which would act as a proxy to CentralNic’s UK-based primary EPP system.
ZDNS, which is prominently promoting its gateway service on its web site, would handle the Whois verification and also proxy the .xyz Whois lookup service, but only as it pertains to Chinese registrants and queries originating in China.
Data on non-Chinese registrants would continue to be housed with CentralNic.
ZDNS would also prevent Chinese registrants registering domains containing strings that have been banned by the government.
XYZ’s RSEP request (pdf) is currently undergoing its technical/competition review with ICANN. Assuming it passes, it would be exposed to public comment before being approved.
The RSEP states: “we are confident that the entire Internet user base of China would endorse this service and that Chinese registrars would strongly endorse this service.”
It’s the first such request to ICANN, suggesting that an awful lot of gTLDs are still not compliant with the Chinese regulations.
As of April, only 14 TLDs — all managed by China-based companies — were licensed to operate in China.

Two .cpa applicants lose CPE

Two applicants that applied for the gTLD .cpa as a “Community” have lost their Community Priority Evaluations.
The American Institute of Certified Public Accountants scored 11 points out of 16, CPA Australia scored 12.
While relatively high scores for CPE, they both failed to pass the 14-point winning threshold.
The string, which stands for “certified public accountant”, is contested by a total of six applicants, which will now have to fight it out at auction.
Both applicants failed to score any of the four available points on the “nexus” criteria, which require the applicant-defined community to closely match the community described by the string.
In both cases, the CPE panel noted that the applicant wanted to restrict .cpa to members of their organizations, which only represents a subset of CPAs in the world.
The decisions can be found here.
Only two CPEs now remain unresolved — the reevaluation of DotGay’s .gay, and DotMusic’s .music. The status of .med and .kids is currently unknown.