Latest news of the domain name industry

Recent Posts

ShortDot plans domain blocking service for brands

Acquisitive new gTLD registry ShortDot is planning a trademark-blocking service along the same lines as those offered by some of its rivals.

The company has applied to ICANN for permission to start a new service called ShortBlock, which it compared to blocking offerings such as Donuts’ Domain Protected Marks List.

Such services block trademarked strings from being registered across a portfolio of gTLDs for a price lower than would be charged for individual defensive registrations.

ShortBlock would also permit typo-blocking, ShortDot’s request states.

Similar services are offered by MMX and .CLUB, both of which will shortly be part of GoDaddy.

ShortDot currently has five gTLDs in its portfolio: .cyou, .sbs, .icu, .bond and .cfd.

Its back-end provider is CentralNic. It says it plans to launch ShortBlock just as soon as ICANN approves its Registry Services Evaluation Process request.

Donuts offers name spinner to show potential attacks

Kevin Murphy, May 13, 2021, Domain Tech

Donuts has launched a tool to show off its TrueName offering, which blocks potential phishing attacks at the domain registry level.

It’s like a regular name spinner, but instead of showing you available domains it shows you visually confusingly similar domains — homographs — that it will block if you register said name in any of Donuts’ portfolio of 2xx (subs, please check) TLDs.

For example, spinning truename.domains returns results such as trʋenɑme.domains (xn--trenme-exc57b.domains) and trᵫname.domains (xn--trname-xk6b.domains), which could be used in phishing attacks.

How many strings get blocked depends largely on what characters are in your name. The letters I and O have a great many visually confusing variants in other non-Latin scripts, and each instance exponentially increases the potential attack vectors.

For example, if I were to register “domainincite” in one of Donuts’ TLDs, Donuts would block 767 homographs at the registry level, but if I were to register “kevinmurphy”, it would only need to block 119.

It only blocks the homographs in the same TLD as the original name. It’s not a replacement for brand protection in other TLDs.

Donuts doesn’t charge anything extra for this service. It’s included in the price of registration and offered as a unique perk for Donuts’ selection of gTLDs.

I gave TrueName a brief post when it launched last year, but I have to say I really like the idea. It’s a rare example of true innovation, rather than simple money-grubbing, that has come from the new gTLD program.

If Verisign were to roll out something similar in .com, it would eliminate a bunch of phishing and cut down on legal fees for big brands chasing phishers and typosquatters through UDRP or the courts.

It was born out of Donuts’ Domain Protected Marks List product, which allows trademark owners to block their brands and homographs across the whole Donuts stable for less money than defensively registering the names individually.

The downside of the spinner tool is of course that, if you’re a bad guy, it simplifies the process of generating samples of homograph Punycode (the ASCII “xn--” string) that can be used in any non-Donuts TLD that supports internationalized domain names.

The tool is limited to 10 domains per spin, however, which limits the potential harm.

Try it out here.

$40 million UNR auction brings fresh blood to domain industry

Six entities are entering the domain registry business for the first time following UNR’s auction last month, which saw over 20 new gTLDs sold off for a total of over $40 million, according to UNR.

While playing its cards close to its chest and revealing the auction results in rather general terms, UNR disclosed last week that there were 17 bidders at the three-day event, which ran in late April.

It said “between 10 and 20 bidders came away as winners”, which I assume we have to interpret as “between 10 and 17”.

Anyone predicting a bulk purchase by a rival portfolio registry was dead wrong, it appears.

UNR said that, while it will not disclose their identities, “established registries, investment firms, blockchain companies, and high net-worth individuals” were among the winners.

None of the ICANN Registry Agreements have yet changed hands, according to ICANN records.

While existing registries and investment firms (presumably the kind of private equity interests that have shown high levels of interest in the domain industry in recent years) will come as no surprise as buyers, blockchain companies and high net-worth individuals will perhaps raise more eyebrows.

ICANN won’t, to the best of my knowledge, sign an RA with an individual, so we’ll no doubt be seeing a corporate vehicle or two established to take over contracts on behalf of those buyers.

The idea of a blockchain company taking over a TLD in the internet’s official root zone is particularly interesting.

The closest we’ve had to that scenario to date is MMX’s experiments integrating .luxe into the Ethereum blockchain, which has been described as genuinely innovative.

But most forays by blockchain outfits into “domain names” have been strictly alt-root moves, such as Unstoppable Domains’ use of .crypto addresses, which do not use the ICANN root and instead require browser plug-ins to function.

These kinds of services usually have their ability to avoid centralized oversight and control as a USP, which makes an attempt from this sector to suck on the ICANN teat especially intriguing.

And which of UNR’s TLDs would be most suited to blockchain applications? .link? .click? .lol?

UNR has not broken down how much was paid for each TLD, and we’ll likely never know, but the $40 million top-line is far above the $11.65 million minimum opening bids it had established for the no-reserve auction.

But it still works out as under $2 million on average across each of the 23 gTLDs on offer, many of which had been on the market for six or seven years, begging the question of whether UNR CEO Frank Schilling’s big bet on new gTLDs back in 2012 was ultimately a success.

Schilling said in a press release: “All UNR shareholders should be exceptionally pleased with the final outcome of this first-of-its-kind event. We are deeply satisfied to have seen so much new interest and blood enter the arena.”

The TLDs auctioned were: .audio, .blackfriday, .christmas, .click, .country, .diet, .flowers, .game, ,guitars, .help, .hiphop, .hiv, .hosting, .juegos, .link, .llp, .lol, .mom, .photo, .pics, .property, .sexy and .tattoo.

DI will of course reveal the winners over time as their ICANN contracts are updated to reflect the new operators.

Kiwis finally making the switch to EPP with Canadian deal

Kevin Murphy, April 19, 2021, Domain Registries

InternetNZ has picked the winner in its registry replacement project RFP, which will see it switch the entire .nz back-end to the industry standard EPP protocol by the end of next year.

It’s selected the CIRA Registry Platform from Canadian .ca registry CIRA, but will continue to run its own back-end in-house in New Zealand.

InternetNZ had said last October that it planned to overhaul its outdated infrastructure, and put out its feelers for would-be vendors or service providers.

At that time, 65% of its over 720,000 .nz registrations and about 65% of its registrars were still using its old, proprietary Shared Registration System protocol.

Now, SRS is to be deprecated in favor of the CIRA platform and the Extensible Provisioning Protocol that has been industry standard across all gTLDs and many ccTLDs for the better part of two decades.

And they say New Zealand is progressive.

InternetNZ plans to make the switch fully by the end of next year. This is of course going to require some implementation work by registrars, which will have to code new hooks into the .nz registry.

UPDATE: This article was updated April 20 to correct “this year” to “next year”. InternetNZ plans to finish the switch before the end of 2022, not 2021.

Verisign says it needs .web because .com is running out of names

Kevin Murphy, April 14, 2021, Domain Registries

Verisign’s affinity for cognitive dissonance has emerged yet again — it’s now claiming that it needs to be awarded the .web gTLD because it’s running out of .com domains to sell.

In legal documents released by ICANN yesterday, Verisign’s lawyers say: “The undisputed evidence is that Verisign needs a TLD like .WEB for growth given the decreased name availability in .COM”.

The admission/claim/lie (delete according to preference) came in a joint post-hearing filing by Verisign and Nu Dot Co, the .web applicant to which Verisign loaned $135 million to bid for the gTLD on its behalf at a record-breaking ICANN auction in 2016.

Afilias, now owned by Donuts, was the second-highest bidder and since November 2018 has been trying to get the auction result cancelled via ICANN’s quasi-judicial Independent Review Process.

The IRP’s final hearing was held over seven days last June, and we’ve been waiting with baited breath for a ruling ever since.

At some point over the last 48 hours, ICANN published three sets of post-hearing arguments — one from itself, one from complainant Afilias and an amicus (non-party, friend of the court) filing from Verisign/NDC.

The Verisign filing (pdf) attempts to rubbish Afilias’ claims across the board, but its rebuttal of the argument that it only wants .web in order to bury it and protect .com’s dominance is particularly interesting:

Verisign Has Every Incentive To Grow .WEB Aggressively. Afilias’ Amended IRP Request asserts without evidence that Verisign seeks to acquire .WEB in order to eliminate a potential competitor for .COM and that Afilias would make a better operator of .WEB. Afilias presented no evidence to support this claim prior to the IRP, and none was presented at the hearing. In fact, the evidence before this Panel refutes Afilias’ claims. The undisputed evidence is that Verisign needs a TLD like .WEB for growth given the decreased name availability in .COM. Even Afilias’ own experts concede that the .COM TLD now has limited name availability. Moreover, the undisputed evidence establishes that Verisign is well-positioned to maximize .WEB’s potential, while Afilias’ recent track record suggests that it would be a less effective operator of .WEB.

In June last year, Verisign had submitted to the IRP panel:

Verisign needs a new TLD like .WEB for growth. Verisign’s growth rate has declined in recent years, largely due to many names in .COM already having been taken and increased competition from new gTLDs and ccTLDs that have superior name availability.

Even Afilias’ own experts concede that the .COM name space effectively is taken. Numerous other industry participants have noted that most of the “good” names in .COM already are taken.

While Verisign had a applied for a few non-English transliterations of .com in the 2012 new gTLD application round, it had avoided getting involved with potential competitors to .com.

But, according to its brief, in 2014 it had just sold off the remainder of its non-domain businesses and, realizing its growth now needed to come from a pure domains strategy, tasked VP Paul Livesay with figuring out how it could worm its way back into the new gTLD program.

Many of the details of Livesay’s research and decision making have been redacted by ICANN (purportedly at Verisign’s request), but it seems he came to the conclusion that the best way to benefit from the program long after the application window closed would be to secretly financially back NDC’s participation in the .web auction, with the provision that the .web contract would be transferred to Verisign should it win.

Quite apart from its regular postings touting .com availability over the last few years, the same year that Verisign was coming to the conclusion that .com was becoming saturated and it needed new growth opportunities in other TLDs, it sued XYZ.com for false advertising for having the gall to suggest that it was hard to find available .com domains. It lost.

Because Verisign apparently enjoys nothing more than holding two diametrically opposed positions simultaneously, its October amicus filing also claims that .web isn’t nearly as awesome as Afilias and others claim.

On the same page that it insists that .web is needed to drive growth, Verisign poo-poos the notion that .web could be a significant competitor to .com, relying on an “expert report” commissioned by Verisign and compiled by University of Chicago economist Kevin Murphy.

(Murphy’s report is redacted in its entirety (pdf) by ICANN, but his 1,119 pages of unredacted exhibits (pdf) prominently include screenshots from this blog, so I feel the need to point out that he’s a different Kevin Murphy — he’s not me, and I’d never even heard of the dude until this morning. On a personal level, the fact that I’m apparently not even the best Kevin Murphy when it comes to the .web story that I’ve been covering for the last two decades is, as you might imagine, as depressing to me as it is presumably amusing to you.)

While his report is redacted, reading around the edges it appears that Murphy reckons .web will not be an exceptional competitor to .com.

Verisign’s October filing states:

.WEB’s Valuation Shows It is Not Particularly Competitively Significant. The Murphy Report models multiple economic scenarios to assess Afilias’ claim that the $135 million price paid for .WEB at the public auction shows that .WEB will be a substantial competitor. None of these scenarios indicate that .WEB is likely to gain a significant market share. Instead, each scenario shows that .WEB is likely to have no more than a 2–3% market share.

Because of the redactions, it’s not clear what market Murphy was referring to, but a 3% market share of the current universe of domain names across all TLDs works out to over 10 million domains. In other words, .web could be a top-five gTLD, up alongside the likes of .org.

But elsewhere in its IRP filings, Verisign cites Murphy to support its argument that .web will have “registrations in the low single digit millions”. That would still be enough to make it one of the best-selling new gTLDs.

This relatively low expected turnout of course begs the question of why Verisign needs .web to grow. It added 4 million net new names across .com and .net last year alone, with .net pretty static, according to its financial filings.

I’m no Kevin Murphy, but here’s a table I’ve thrown together showing Verisign’s domain growth over the last decade.

[table “64” not found /]

Its revenue has consistently grown year over year, from $681 million in 2010 to $1.27 billion in 2020. It’s considered one of the most profitable companies in the world, and its share price has tripled since 2011.

And that was without .web.

EFF rages as Ethos closes Donuts buy

The Electronic Frontier Foundation thinks the acquisition of Donuts by “secretive” private equity group Ethos Capital represents a risk to free speech.

The deal, which sees Ethos buy a controlling stake from fellow PE firm Abry Partners, closed earlier this week, having apparently received no official objection from ICANN.

But the EFF now wants ICANN to force Donuts to change its gTLD registry contracts to make it harder for the company to engage in what it calls “censorship-for-profit”.

The group’s senior staff attorney, Mitch Stoltz, raised the issued at the Public Forum session of last week’s ICANN 70 virtual public meeting, and expanded upon his thinking in a blog post this week. He wrote:

Donuts already has questionable practices when it comes to safeguarding its users’ speech rights. Its contracts with ICANN contain unusual provisions that give Donuts an unreviewable and effectively unlimited right to suspend domain names—causing websites and other internet services to disappear.

He pointed to Donuts’ trusted notifier program with the Motion Picture Association, which streamlines the takedown of domains used for pirating movies, as an example of a registry’s power to censor.

Donuts runs gTLDs including ones with social benefit meanings that the EFF is particularly concerned about, such as .charity, .community, .fund, .healthcare, .news, and .university.

Stoltz also makes reference to the Domain Protected Marks List, a Donuts service that enables trademark owners to block their marks, and variants, across its entire portfolio of 240+ gTLDs.

In effect, this lets trademark holders “own” words and prevent others from using them as domain names, even in top-level domains that have nothing to do with the products or services for which a trademark is used. It’s a legal entitlement that isn’t part of any country’s trademark law, and it was considered and rejected by ICANN’s multistakeholder policy-making community.

The DPML is not unique to Donuts. Competitors such as UNR and MMX have similar services on the market for their gTLDs.

When Stoltz raised the EFF’s concerns at last week’s ICANN meeting, CEO Göran Marby basically shrugged them off, saying he didn’t understand why one PE firm buying an asset off another PE firm was such a big deal.

I have to say I agree with him.

Ethos came under a lot of scrutiny last year when it tried to buy .org manager Public Interest Registry, turning it into a for-profit entity, generating cash for Ethos’ still-undisclosed backers.

(This week, Ethos disclosed in a press release that its investors include massive hedge funds The Baupost Group and Neuberger Berman “among others”, which appears to be the first time these names have been mentioned in connection with the company).

But a pretty good case could be made that .org is a unique case, that has had a non-profit motive baked into its DNA for decades. That does not apply to Donuts, which was a profit-making venture from the outset.

It’s not entirely clear why the EFF is suddenly concerned that Donuts will start exercise its contractual right-to-suspend more frequently under Ethos than under Abry. Stoltz wrote:

As we learned last year during the fight for .ORG, Ethos expects to deliver high returns to its investors while preserving its ability to change the rules for domain name registrants, potentially in harmful ways. Ethos refused meaningful dialogue with domain name users, instead proposing an illusion of public oversight and promoting it with a slick public relations campaign. And private equity investors have a sordid record of buying up vital institutions like hospitals, burdening them with debt, and leaving them financially shaky or even insolvent.

Even with the acquisition passing through ICANN easily, the EFF wants Donuts to change its contracts to make it more difficult for the company to suspend domain names on a whim.

I believe the language causing the controversy comes from anti-abuse policies in the Public Interest Commitments found in almost all Donuts’ contracts with ICANN, which state in part:

Registry Operator reserves the right, at its sole discretion and at any time and without limitation, to deny, suspend, cancel, or transfer any registration or transaction, or place any domain name(s) on registry lock, hold, or similar status as it determines necessary for any of the following reasons:

a. to protect the integrity and stability of the registry;

b. to comply with any applicable laws, government rules or requirements, requests of law enforcement, or any dispute resolution process;

c. to comply with the terms of this Registry Agreement and the Registry Operator’s Anti-Abuse Policy;

d. registrant fails to keep Whois information accurate and up-to-date;

d. domain name use violates the Registry Operator’s acceptable use policies, or a third party’s rights or acceptable use policies, including but not limited to the infringement of any copyright or trademark; or

e. as needed during resolution of a dispute.

As a voluntary PIC, this language is unique to Donuts, though other registries have similar provisions in their registry agreements.

Universal Acceptance – making the internet work for everyone [Guest Post]

Kevin Murphy, March 24, 2021, Domain Tech

Editor’s note: this is a guest post written by Aman Masjide, head of compliance at new gTLD registry Radix.

Back in 2014, to foster innovation and to better the choice in domain names, ICANN introduced new generic top-level domains through its New gTLD Program. It was a monumental move that enabled businesses, individuals, and communities across the globe to mark their presence on the internet.

Allowing users to be present digitally in their chosen language (non-ASCII characters and scripts) gave opportunities to local businesses, civil societies, and governments to better serve their communities.

Analysys Mason conservatively estimates that there is scope of $9.8 billion growth in potential revenue from both; existing users who are using new domain names and from new internet users coming online through Internationalized Domain Names (IDNs).

To achieve this, Universal Acceptance of new gTLDs and IDNs is critical in making the Internet more accessible to the next billion users. Founded in February 2015, the Universal Acceptance Steering Group (UASG) undertakes activities to promote Universal Acceptance of all valid domain names and email addresses.

Through its ambassadorship and local Initiative programs, UASG promotes Universal Acceptance globally. Their efforts are divided and executed through five working groups that include:

  • Technology Working Group
  • Email Address Internationalization Working Group
  • Communications Working Group
  • Measurement Working Group
  • Local Initiatives Working Group

Before we get into the acceptance of new domain extensions (nTLDs), we must first understand what acceptance means and how it’s measured.

The Universal Acceptance Steering Group’s mission sums up acceptance in one short statement: “All domain names and all email addresses work in all software applications.”

While this is a simple understanding of the concept, for an end user of an nTLD, this statement further branches out into multiple questions such as:

  • Will my domain name work on all platforms/applications–online or offline?
  • Will my email address on a new domain extension get accepted on all websites/platforms and pass all the validation tests?
  • Will my emails on new domain extensions, once accepted, stop going into the junk folder?
  • Will I be able to use all the features of a website/platform irrespective of my domain extensions? For example, will a social media platform accept a new domain extension in the bio, comments, posts, messenger, etc, and process it exactly like any other legacy TLD?

The Universal Acceptance (UA) of all domain names and email addresses requires that every piece of software is able to accept, validate, process, store, and display them correctly and consistently.

As a new domains registry, it was critical for us to understand what the gaps were and how to close them so that the internet operates the same for nTLD users as it does for the legacy TLD users.

Initial research concluded that UA readiness issues occur when applications are not able to handle the following categories of a domains name or email addresses:

Domain Names

  • New short top-level domain names: example.fun, example.site
  • New long top-level domain names: example.berlin, example.space
  • Internationalized Domain Names: παράδειγμα.ευ

Email Addresses

  • ASCII@ASCII; new short or long TLD: ekrem@misal.istanbul
  • ASCII@IDN: john@société.org
  • Unicode@ASCII: 测试@example.com
  • Unicode@IDN: ईमेल@उदाहरण.भारत
  • Unicode@IDN; right to left scripts: لیم@لاثم.عقوم ای

For Universal Acceptance to succeed, it needs to be examined holistically.

Over the years, UASG working group members have conducted several gap analysis on programming languages and frameworks, networking command-line tools, web browsers, websites, and have made great strides in acceptance of new domain extensions.

According to UASG’s FY 2020 report, tests conducted on top websites showed that

  • The acceptance rate of emails on short nTLDs has increased from 91% in 2017 to 98.3% in 2020.
  • The acceptance rate of emails on long nTLDs has increased from 78% in 2017 to 84.8% in 2020.

table

Note: The table above compares the 2020 results to the earlier 2017 and 2019 testing results.

Two important caveats should be remembered in this case:

  • Different email addresses were tested (but they were of the same type).
  • The websites tested in 2020 were different from previous ones as they were the 50 most popular in the 20 countries rather than the 1,000 most popular globally.

However, these results may still be used to compare overall trends.

Universal Acceptance Readiness Report 2020 (pdf) also segregated test websites as per different categories such as eCommerce, government, education, etc and the results were promising.

table

Such studies help UASG ambassadors and advocates to identify and focus on websites of a specific category that require immediate attention. We conducted a similar study at Radix where we analysed top websites belonging to different categories. These were the results (click to enlarge):

table

While the acceptance rates for new short and new long cases is more than 80% under most categories, we see a drastic dip when a domain is on an IDN TLD. Such comparisons highlight problem areas and provide direction to ambassadors and members who are advocating for Universal Acceptance.

Radix’s contribution to UASG

UA is something that affects nTLD users the most. This is why it’s crucial to focus on the feedback that we receive from them. At Radix, we work closely with our users to ensure we have the first hand information on any UA related issues faced by the customer.

The feedback could be about linkification, validation or acceptance of emails on nTLDs on different websites and platforms. Radix also actively invests its resources in gap analysis by testing various websites and social media platforms. We are also part of the ambassadorship program promoting and supporting local and global UA initiatives.

Here are some of the UASG initiatives that Radix is part of:

At Radix, our objective is to ensure that nTLDs are accepted across websites and platforms. To achieve this, we actively work with UASG and share as many issues and gaps noticed and reported by customers.

Contribution by other registries

A key objective for most registries is to ensure great customer experience when it comes to their nTLDs and I’ve always admired it when registry operators have actively taken initiative and participated in the five UASG groups mentioned above.

One of the ways to do this is to capture all the queries and complaints reported by their customers/registrar partners and share it with UASG. This will help their support team direct their resources in solving the problems and encouraging those websites to become UA compliant.

Contribution by registrars

When it comes to UA-related issues, registrars are the first in chain to receive a complaint or feedback from the user. Therefore, it’s crucial that their support teams have all the necessary information needed on how to best handle such complaints.

For now, they can:

  • Inform the customer about the potential UA issue and raise a request on behalf of the customer with UASG. Issues can be logged at – https://uasg.tech/global-support-center/
  • Report these instances to the Registry Operator so that they can connect and follow up with UASG.
  • Join any of the five working groups and participate.

The path ahead

The UASG is consistently compiling and sharing all the important information needed for organizations and developers to become UA ready. This is not only about ensuring the readiness of a system to accept certain TLDs or emails, but also about realising the full potential of an organization by connecting with people and businesses that might not be even on it’s radar.

Every successful step taken by an organization towards UA readiness is also a step towards equality and inclusiveness on the internet.

Guest poster Aman Masjide leads compliance and abuse mitigation at Radix.

Motion to fire five Nominet directors passes in tight vote

Kevin Murphy, March 22, 2021, Domain Registries

Nominet’s members this afternoon voted to fire five of the .uk registry’s directors, including its chair and CEO, in an unexpectedly tight poll.

There were 2,432,105 votes in favor of the motion to fire Mark Wood, Eleanor Bradley, Ben Hill and Jane Tozer, compared to 2,179,477 against, which works out to 52.74% for and 47.26% against.

Members get allotted votes according to how many domains under management they have, capped at 3% of the total cross-membership vote count.

In the end, it appears that the vote was swung by a small handful of larger registrars. Tucows and Namecheap were the largest registrars to say they would vote for the board cull.

Turnout was 53.5% of eligible voters, which I gather is extraordinarily high for a Nominet member vote. The full results are here (pdf).

The original motion also named CEO Russell Haworth, but he quit his board seat and executive role yesterday.

Bradley and Hill, respectively managing director of the registry and CFO, have left the board but keep their staff jobs.

Rob Binns is now acting chair, Nominet said.

The vote took place at an Extraordinary General Meeting held virtually this afternoon, which was called for after 5% of Nominet’s registrar/domainer members signed a petition at PublicBenefit.uk.

The campaign was orchestrated by Simon Blacker of the registrar Krystal Hosting, reflecting growing displeasure among members about Nominet’s strategic direction and lack of member engagement.

After the result was announced this hour, Blackler was quick to hail it as “a watershed moment in Nominet’s history”, saying that it “demonstrates the resolve of the membership to restore its original purpose”, in a letter (pdf) to the “remnant” board.

He went on to call for the campaign’s original two picks for chair and vice-chair replacements — Sir Michael Lyons and Axel Pawlik — to be appointed to the board on an interim basis.

The EGM, which lasted for about an hour, saw directors repeatedly acknowledge and occasionally apologize for taking too long to recognize the breadth and depth of members’ concerns, promising to turn things around.

The key theme to emerge was that the company is now on the same page as its members and is committed to addressing their concerns, but that eliminating almost half of the 11-person board would delay these actions by months.

Wood also reiterated that the threat of UK government intervention is real, should it be perceived that Nominet — a piece of critical infrastructure in all but name — was becoming unstable.

The EGM result was due around 1800 UTC but was delayed by more than three hours, apparently due to the higher than expected turnout.

Nominet warns of government takeover as Namecheap backs fire-the-directors campaign

Kevin Murphy, March 15, 2021, Domain Policy

Nominet has raised the specter of a government takeover of the .uk registry, should members vote to oust five of its top directors at an Emergency General Meeting a week from now.

The warning came as part of the company’s anti-EGM publicity, and at a time when the campaign for a Yes vote has passed 25% of eligible votes, with Namecheap becoming the biggest name yet to support the ouster.

In a blog post, the company refers readers back to the Digital Economy Act of 2010, which in part gives the UK government the ability to unilaterally take over .uk, should Nominet seriously mess up:

It means that the government can step in, if Nominet is ever considered unstable or not capable of governing itself.

Removing five directors, including two of the four independents, and pressuring the remaining directors to install candidates outside normal procedures, as the EGM petitioners seek to do, would be a huge step backwards in terms of good governance. We have been warned that instability will be of serious concern to government. We know it would create a scenario which would make intervention more likely.

That part of the Act was brought in because at the time Nominet was perceived to be at risk of capture by domain investors. It has since reformed its constitution to make this less likely.

The current situation could be seen as a replay of the situation 11 years ago, with many of those most unhappy with Nominet’s recent strategy among the domainer community.

The campaign, PublicBenefit.uk, wants to fire five directors including the chair and CEO and replace them with two new appointments who have promised to lower .uk domain prices and direct more profit to public benefit causes.

As of today, the campaign has 429 member signatories, representing 25.1% of voting rights. This is probably enough to pass its resolutions, which call for a simple majority of members attending the EGM.

Namecheap has become the largest registrar so far to sign up. It’s the seventh-largest .uk registrar, with 201,355 domains under management. GoDaddy, 1&1 Ionos, Tucows, and the others in the top 10 are so-far undecided. Google has said it will abstain.

It’s debatable whether the Digital Economy Act applies here. The Act deems that a registry has failed under two quite narrow circumstances:

(a)the registry, or any of its registrars or end-users, engages in prescribed practices that are unfair or involve the misuse of internet domain names, or

(b)the arrangements made by the registry for dealing with complaints in connection with internet domain names do not comply with prescribed requirements.

Do either of those apply to PublicBenefit.uk’s demands? It looks like a stretch.

The EGM will take place March 22, next Monday, and right now it’s not looking great for Nominet’s top brass.

Domain industry shrank in Q4, but as usual there’s a big BUT

The worldwide domain name count shrank in the fourth quarter, according to newly released Verisign data, but as usual the numbers were hugely impacted by big swings in just a few TLDs.

The latest Domain Name Industry Brief (pdf), which is mainly compiled from zone file counts, shows that 2020 ended with 366.3 million names, down by 4.4 million or 1.2% compared to the end of the third quarter.

It’s the free and almost-free TLDs that swung the math.

Remarkably, industry wild-card .tk actually shrank during the quarter. This is highly unusual, as the registry’s business model is based on giving out names for free, never deleting domains, and monetizing the traffic to expired or suspended names.

It saw domains down by 2.8 million names over the quarter, from 27.5 million to 24.7 million.

Another big dipper was .icu, which sells cheap (usually under $1) and appeals to speculators largely in China.

While it slipped out of the top 10 TLDs, meaning the DNIB no longer breaks out its numbers, DI’s own zone file counts show its zone decline from 5.3 million to 3.4 million during Q4, a 1.9 million decline.

Notably spammy new gTLD .top, which also costs next to nothing and is popular in China, also had a role to play. Its zone count was down by about 900,000 between September 30 and December 31.

Those three TLDs alone account for a loss of 5.6 million names, far more than the 4.4 million industry-wide quarterly drop calculated by Verisign.

The impact of .icu’s continued spiral downwards is likely to be felt in Q1 2021 also. It’s lost another 2.4 million zone file names since the start of the year.

Verisign said the the universe of ccTLD domains contracted by 1.7 million of 1% during the quarter, ending the year with 158.9 million names.

The .tk shrinkage of course more than accounts for this dip. Without it, ccTLDs would be up by 1.1 million names or 1.1%. The major, top-10 ccTLDs mostly showed six-figure growth, the DNIB reflects.

New gTLDs were down 4.2 million names or 13.8% sequentially, ending the quarter with 26 million.

In addition to the aforementioned .top and .icu, this figure appears to have been affected by six-figure losses in some of the highest-volume, lowest-priced new gTLDs, including .club, .site .work and .vip.

In the main legacy gTLDs, Verisign’s own .com grew by 1.5 million names, from 151.8 million to 150.3 million, during the quarter. Its .net was again flat at 13.4 million. Public Interest Registry’s .org gained a (rounded) 100,000 names, ending the year at 10.3 million.

The annual numbers across the industry for 2020 have better optics. The DNIB shows that domain volume was up by 4.0 million or 1.1% year over year.

That breaks down into a 6.3 million increase in .com, a 1.3 million increase across the ccTLDs, and a 3.3 million decrease in new gTLDs, not all of which can be explained away by factoring out .icu and .top.