[CPWG] To convey, not to judge

[John McCormac]

On 06/05/2020 16:53, Evan Leibovitch wrote:
> 
> When ALAC expresses an opinion that's unpopular with the domain industry 
> we are invariably greeted by "who the hell are you and what gives you 
> the right to speak on behalf of the billions of the world's Internet 
> users?" On reflection, they're right. We have about 25 At-Large leaders 
> (give or take a few liaisons) attending each ICANN meeting, elected or 
> appointed by RALOs and other self-selected volunteers, who really have 
> absolutely no authority or rationale to act as a filter for the public 
> sentiment. We keep busy by chasing ICANN public comments and the agendas 
> of others, carefully crafting commentary on things that mean a great 
> deal to the domain industry but not a shred to the outside world. And by 
> and large we avoid issues that would collide with ICANN's 
> highly-corrupted goals, so as not to incur the "who the hell are you?" 
> retort.

If ALAC doesn't speak up then who will?

> How to fix? Needs to start from the ground up, bandaids won't fix this 
> mess. ICANN At-Large needs to be designed to highlight researchers and 
> writers, not petty politicians and lobbyists. I would personally trade 

So does that mean that lobbyists should be red flagged as being such?

> almost every dollar spent on ALAC attending ICANN meetings for a decent 
> R&D budget and the ability to do global surveys. How many people here 

And produce the a kind of CCT product filled with great aspirations and 
good intentions? There was no expertise on measuring web usage or other 
important industry metrics on that team. The CCT web usage measurement 
attempt (don't think the attempt at measuring web usage made it to the 
final report) gave the new gTLDs a glowing report.

The reality was different. And within about a year, many of the domain 
names in these wonderfully used new gTLDs were deleted. Webspam and 
parked pages were considered as being actively used. Some new gTLDs were 
doing relatively well. The problem was that ICANN and the CCT did not 
understand how TLDs develop or how domain names are used. ICANN, being 
the customer for the CCT report, had a flawed view of the new gTLD 
market's development.

As for global surveys, they sound nice and very impressive. The problem 
is that the domain name market has a small global market and many 
country level markets. In many of those country level markets the gTLDs 
are only a second choice for new domain name registrants. As for Africa 
and Latin America, there are hardly any ICANN accredited registrars in 
those regions but many of those countries have strong local ccTLDs. The 
domain name world changed while ICANN wasn't looking.

The opinion poll surveys commissioned by the CCT to see if people had 
heard of the new gTLDs were even funnier though they were 
methodologically sound as opinion polls. Some parts of the CCT report 
were on much firmer ground but the new gTLDs did not turn out as ICANN 
expected. The people in the domain name industry, the ordinary domain 
name registrants and people on the street knew that well before ICANN.

As for ICANN's expectation of over 30 million registrations in the first 
year, I really wonder about how they come up with those figures but it 
must have involved some psychedelic grade ICANN Koolaid. No wonder those 
ICANN meetings are so popular.

How would this R&D and these global surveys be any different?

> The UN-like meetings and fake self-importance need to stop. Our main job 
> is to listen to and understand, not act as a gatekeeper for, what the 
> world is saying on DNS issues. Our prime task is to make sure that ICANN 
> is aware of the public PoV, as accurately possible, regarding:
> 
>   * The (to many needless and abuse-generating) expansion of TLDs

 From what I've seen, and I actually run web usage surveys on gTLDs and 
ccTLDs, the majority of abuse occurs when the business model of a gTLD 
fails and the registry has to adopt discounting as a business model in 
order to survive. That brings in tens of thousands or even millions of 
low quality registations (robot registrations where the domain name is 
generated by an algorithm) filled with webspam and sites built from 
scraped content. These domain names have a renewal rate of under 5% so 
the registry has to keep discounting to replace the deleting domain 
names with new registrations. Some new gTLDs are being used for websites 
and e-mail. The credibility of the TLD also collapses and new 
development and usage in the TLD ceases.

>   * The ability of registrants to assert rights protections in domain
>     space that are disallowed globally for trademarks
>   * The utility of "memorable" domain names versus search engines and
>     social media
>   * The practise of domain hoarding and speculation (also illegal under
>     international trademark regime)

How can this "international trademark regime" decide what is legal and 
what is illegal? Are they like the Spanish Inquisition?

This is a highly problematic issue in that a lot of the assumptions 
about what is "hoarded" or speculated is wrong. The logic generally 
seems to be that because someone registered a domain name first and the 
domain name has no active website, everyone else who wants that domain 
name thinks that the registrant is hoarding it.

In the last fifteen years or so, registrars have been parking 
undeveloped domain names on PPC services. So if someone sees a domain 
name on PPC, the assumption is often that it is "hoarded". The reality 
is that the majority of domain names in both gTLDs and ccTLDs have no 
developed websites. The first year renewal rate for .COM is 
approximately 57%. In 2004, it was running at around 70%. The registries 
generally publish blended renewal rates that represent all renewed 
domain names rather than first year renewals.

The reality, as demonstated by the 01 May 2020 stats for .COM is that 
5,888,117 domain names (approximately 4% of .COM) are on major 
auction/sales sites. There are other smaller auction sites and DIY For 
Sale sites and webpages that are not included in this figure. Some of 
these domain names have been registered for decades or have been put on 
these sites by their registrants. Others have been reregistered when 
they are deleted. Then there are the sales sites where people come up 
with "brandable" domain names. The secondary market is complex. The .ORG 
percentage was 2.4%.

>   * Whether domain names need regulation or just the basic sanity checks
>     now on offer

Regulation? What people inside the ICANN bubble don't realise is that 
apart from the USA where the .COM is the de facto ccTLD the domain name 
the market is shifting away from gTLDs to ccTLDs. In many countries 
where there is a strong ccTLD, growth in the gTLDs has plateaued with 
the main volume of new registrations being in the local ccTLD. Some of 
the legacy gTLDs in these markets are existing only on brand protection 
registrations. Regulation, in a wider context, is being gradually 
removed from ICANN's hands. This isn't a kind of Balkanisation of gTLDs 
and ccTLDs. It is unexpected obselescence for some gTLDs.

The funniest thing about the rise of the ccTLDs is that it was ICANN's 
attempts to deal with Domain Tasting that accelerated the shift away 
from gTLDs towards ccTLDs. Between 2005 and 2007, over one billion .COM 
domain names were deleted. That's over 1,000,000,000 domain names. The 
monthly numbers are in the free to read section of the Domnomics book on 
Amazon and they constitute a horrifying illustration of a failure to 
regulate. (I think that he ALAC/GNSO paper on Domain Tasting only 
focused on the first month that the AGP deletion figures were made 
available.)

ICANN and the constituencies discussed the matter and even tried to 
figure out what was going on with all these registrations but ICANN and 
the constituencies failed. It took external action to bring the matter 
to a conclusion and that was the Dell law suit. ICANN was plodding along 
with various discussions about how to deal with it and even managed to 
get some new rules to stop the abuse of the AGP. But it was from Domain 
Tasting and the artificial scarcity it created that the idea that there 
was a demand for new gTLDs emerged. The problem was that it was based on 
a misunderstanding of the impact of an artificial scarcity of "good" 
domain names on the domain name market. The new gTLD train was rolling 
off down the tracks by the time that large-scale Domain Tasting was 
ended 2009. The expectation (or Astrological prediction) of over 30 
million registrations in the new gTLDs for the first year of operation 
was just another example of ICANN bubble thinking.

The CA AG intervention was yet another example of the ICANN bubble being 
burst by external action. In some respects, ICANN only evolves only when 
there is external intervention. Even the first new gTLDs in the early 
2000s were a response to a perceived "all of the good names are gone" 
meme from the late 1990s. In early 2000, the DotCOM bubble burst and the 
demand for some of these first new gTLDs disappeared as million of .COM 
and .NET domain names flooded back into the market. It is like the ICANN 
bubble assumes that the domain name market evolves steadily. It does not 
and external forces play a much larger part than the ICANN people (and 
even those in some of the constituencies) realise.

>   * The balance between privacy and law-enforcement access to reduce abuse

Now that's a can of worms! The GDPR might have seemed like a good idea 
to those involved but trying to apply a theoretically perfect solution 
to an imperfect world was bound to fail. It even made a complete mess of 
WHOIS. This is the same tool that largely functioned before the abject 
stupidity of deleting all fields for "privacy". In the interests of 
making things better, things were made magnitudes worse and now they 
even facilitate abuse.

And now GAC even wants to introduce completely clueless "uniform" 
anonymous e-mail addresses. The whole point of anonymity is that it is 
not possible to burn through that anonymity by analysing the algorithm 
used to generate the anonymous addresses and the plaintext of hundreds 
of millions of existing e-mail addresses. A "uniform" system is also 
extremely difficult to introduce across hundreds of TLDs, thousands of 
registrars and hundreds of millions of e-mail addresses. It was like 
whoever came up with that GAC statement didn't understand the concepts 
of basic cryptography or the reality of the real world. It was just an 
excercise in buzzword bingo.

> But look elsewhere in this list and you'll find none of those things 
> being the focus our work. Instead good people are caught up in endless 
> development processes about development processes, success metrics etc. 

That does make it sound like a kind of middle-management Hell. But the 
real problem, when it comes to domain names and their markets, is that 
if you can't define what you are trying to measure, then you won't know 
whether it is a success or failure. That's one of the persistent 
weaknesses of ALAC and ICANN in that both the definitions and the data 
are not readily available.

It isn't just due to organisational information silos that exist in 
large organisations like ICANN. It is often down to data being 
commercially sensitive. What happens is that the gaps are filled by the 
assumptions of well-meaning people (as with the CCT) rather than hard 
data. Those assumptions quickly gain the status of being "metrics" by 
equally unaware and well-meaning ICANN management. It seems more like 
the development of an organised religion than a business with data-free 
assumptions quickly becoming doctrine.

And as for the policy development and processes, there's a great phrase 
from the movie "The Outlaw Josey Wales" that describes it: "endeavour to 
persevere". And it doesn't look like ALAC will be declaring war on ICANN 
any time soon.

Regards...jmcc
-- 
**********************************************************
John McCormac  *  e-mail: jmcc at hosterstats.com
MC2            *  web: http://www.hosterstats.com/
22 Viewmount   *  Domain Registrations Statistics
Waterford      *  Domnomics - the business of domain names
Ireland        *  https://amzn.to/2OPtEIO
IE             *  Skype: hosterstats.com
**********************************************************