[Gnso-newgtld-wg-wt5] Conclusion of CWG-UCTN on 2-character codes

Liz Williams liz.williams at auda.org.au
Fri Jan 19 12:44:14 UTC 2018

Hello everyone

Ann-Cathrin has raised some interesting points which are all mixed together.

1.  ASCII two letter combinations are currently used to define country codes which are commonly used geographic identifiers for ccTLD operators.  It seems we all agree on that if I read the sentiment on the list correctly.  And we would maintain the policy recommendation to use the ISO3166 lists to define what those two character combinations are.

2.  Other combinations of 2 characters could easily be IDN scripts (for example using the two characters for Osaka) would be a 2 character geographic term and there are many other examples in many different scripts.  We need to find consensus on how to treat two characters (IDNs) in the context of our work here.  Should they be treated in the same way as ASCII characters?

3.  And then the much more difficult proposition that a gTLD “should" be 3 characters or more.  Of course, we already have examples where two character country codes (for example, .la, .co & .tv) are used generically to register names which have nothing to do with a geographic term or a country code.  I would have to question the assumption that a gTLD of, say, F1, would be anything other than generic, not confusable with a country code but I do think we need clarification on this interesting point.

4.  And then the most difficult line which is that “gTLDs need to be at least three characters”.  This presumes that we continue with the two character ASCII limitation and that many potential new geographic identifiers (for short hand geo names) should be at least three characters…in ASCII and IDN?  If we follow this logic it means that any new TLD that is a geographic identifier (and then it would be evaluated and implemented as such) could be both a geo name and a generic TLD (for registry agreement purposes with all the conditions that attach to that agreement).

5.  And finally, do we put out of scope number/letter, letter/number and number/number combinations as they may not be geographic identifiers?  That is a question which I think needs further examination quite broadly before we decide if, for the purposes of this group, if it is indeed out of scope.  If it is, then what other work track is dealing with it because the issue is open.

Best wishes.


On 19 Jan 2018, at 12:03 pm, Ann-Cathrin Marcussen <ann-cathrin.marcussen at norid.no<mailto:ann-cathrin.marcussen at norid.no>> wrote:

Hi all,

In my opinion it is outside the scope of WT5 to judge anything else than geographic terms. In terms of 2 character (ie F1 and 3M etc)  those, in my mind, aren’t geo terms and thus out of scope.

I am also of the opinion that we should keep the system that the requirement is that a gTLD needs to be at least 3 characters.


Best regards / Med vennlig hilsen

Ann-Cathrin Marcussen
Head of Legal

7465 Trondheim

St Olavs Plass 2
0165 Oslo

Mobil +47 920 14 282
acm at uninett.no<mailto:acm at uninett.no>
On 18 Jan 2018, at 17:16, Alexander Schubert <alexander at schubert.berlin<mailto:alexander at schubert.berlin>> wrote:

Hi Rosalia,

Do you support to reserve only the 2 LETTER codes, or all ASCII two CHARACTER codes? E.g.:  Would you be OK with a gTLD “.f1” or “.3m”?



From: Gnso-newgtld-wg-wt5 [mailto:gnso-newgtld-wg-wt5-bounces at icann.org] On Behalf Of Rosal?a Morales
Sent: Thursday, January 18, 2018 5:47 PM
Cc: Icann Gnso Newgtld Wg Wt5 <Gnso-newgtld-wg-wt5 at icann.org<mailto:Gnso-newgtld-wg-wt5 at icann.org>>
Subject: Re: [Gnso-newgtld-wg-wt5] Conclusion of CWG-UCTN on 2-character codes

Dear All,

As a member of the already closed CWG-UCTN, I support the consensus that the group came about regarding 2-letter TLDs: reserving 2-letter codes for ccTLDs should be maintained.  The effort and many years of discussion of the CWG-UCTN should serve as a great stepping sone for this working group to build on. Opening this for debate once again, will lead to lengthy discussions and we will be reinventing the wheel once again. I strongly recommend supporting reserving 2-letter codes for ccTLDs as concluded by CWG-UCTN.


On Jan 18, 2018, at 8:43 AM, Timo Võhmar <timo.vohmar at internet.ee<mailto:timo.vohmar at internet.ee>> wrote:

Some may know that I represent a bit different view on things than Alexander

We do not need to do anything ie reserve anything to anyone and least of all does the rfc say anything about that. The RFC sets out closed list of what is generic tld and we know that has not been true for a long time. So what we have here in case of two letters is just an agreement between different communities saying that all two letter combinations are reserved for use under ccTLD policy. We also have renewed this agreement multiple times and there seems to be no strong demand in changing this. But at the same time that does not mean that ccTLD cannot be anything else than some two letter combination.

cc in ccTLD stands for country code. Country code can mean any type of label, name, letter combination that represents some country. Unfortunately we do not have community wide agreement on how to treat 3-letter ISO country codes and country names. There is a demand to make these available for use under ccTLD policy as well as gTLD policy. This need to find a compromise seems to be one of the reasons we have gathered.

Of course the question at hand is wider - anything that represents anything with geographical meaning. And we are here by invitation of gNSO with the new gTLD round in mind but I do believe that it is not possible to change the current status quo disregarding the interests of countries to protect and develop their identity. One way to separate generics from country codes is who runs the delegation and perhaps on what purpose.

And I would once again like to argue that anyone outside our community has or even should have any notion about meaning of generic and country code TLDs in the ICANNs policy context. It's all about what the TLD stands for ie there are few people that associate .tv with Tuvalu nor do I see that there could be anything generic about .usa. So we are speaking here about how these TLDs are delegated, regulated and how much control does registry/government have on how to run the TLD. Noone in the real world will get confused when suddenly there are n letter TLDs given out under ccTLD nor 2 character TLDs under gTLD policy - these are just TLDs.

Lets leave the emotional arguments aside and try to find better solution for geographic terms for the next gTLD round and see if we can find a way to lift the ban from country labels.

Best Regards,
Timo Võhmar
Representing Estonian internet community, government and ccTLD .ee registry

On Thu, Jan 18, 2018 at 8:12 AM, Greg Shatan <gregshatanipc at gmail.com<mailto:gregshatanipc at gmail.com>> wrote:
I don't disagree with the idea that all two-letter combinations should be reserved for potential ccTLD use.  I thought it was worth raising the question on our recent call so that we could actively test the status quo on this point.  (Also, nobody was speaking up....)  It appears that this concept retains broad support.

That said, I would like to take Alexander's invitation to correct him if his statements are factually incorrect or illogical.  Specifically, it is both factually incorrect and illogical to couch this in terms of "brands."  There's absolutely no basis or evidence for the proposition that brands have a particular interest in or desire for two letter TLDs (or that other applicants have less or more interest or desire).  Intentionally or not, using that scenario stirs up tensions and creates an adversarial atmosphere that could be quite counterproductive.  There's no reason to invent a disagreement or rift that doesn't exist.

I would therefore suggest that the LAST thing the ccNSO should do is ask themselves the question posed by Alexander above.  There's no good reason to pose that question or to approach this issue in that way, and there are plenty of good reasons not to do so.

Instead, we should approach this as one point on which we have broad support from all sectors, including many from the brand community.


On Wed, Jan 17, 2018 at 10:45 AM, Alexander Schubert <alexander at schubert.berlin<mailto:alexander at schubert.berlin>> wrote:
I think the two character TLD question is rather simplistic:

•         We need to reserve ALL two letter combinations – RFP 1591 says so. Or we need to revise RFC 1591: GOOD LUCK with that one! Easier to revise the Bible – or the Bill of Rights!

•         We can’t assign two number combinations as they could be confused with IP addresses: Example 4th level domain “” based on the gTLD “.11”!

•         What remains are number-letter combinations! That’s 2 times 10 numbers by 26 letters = 520 combinations. With almost NONE of them really desirable for anybody. The rare desirable ones would be likely BRANDS, such as “.3m” (3m.com<http://3m.com/>) or “.f1” (f1.com<http://f1.com/>). How many would that be? A dozen?

On the other hand: Outside of the U.S. (96% of the world population is non-American) EVERYBODY knows that a two letter TLD is “something different”. And that is the one and only distinction the ccNSO world owns: they are the arbiters of the two character namespace on the top-level! Two-characters = ccTLD. Everything else = gTLD.

If I were the ccNSO I would ask myself: Just so that a handful of brands COULD (and nobody knows IF they WOULD want) grab a two character number-letter gTLD: do we have to give up the old order? As ccNSO member and as GAC member I would clearly draw a line here and say: “NO WAY”. 2 characters is the namespace-characterization of the ccNSO (even when they only use two letter combinations). Outside of the U.S. it would very much confuse the Internet User if suddenly SOME 2-character were gTLDs – or 3-letter TLDs were ccTLDs. That would destroy the old order.

If this is factual incorrect or illogical: please correct me. If you support this notion: please voice your support!



From: Gnso-newgtld-wg-wt5 [mailto:gnso-newgtld-wg-wt5-bounces at icann.org<mailto:gnso-newgtld-wg-wt5-bounces at icann.org>] On Behalf Of Heather Forrest
Sent: Wednesday, January 17, 2018 8:42 AM
To: Gnso-newgtld-wg-wt5 at icann.org<mailto:Gnso-newgtld-wg-wt5 at icann.org>
Subject: [Gnso-newgtld-wg-wt5] Conclusion of CWG-UCTN on 2-character codes

Dear colleagues,

Following up on Jeff's reference at the end of the WT5 call ended just now to the conclusions of the Cross-Community WG on the Use of Country and Territory Names (CWG-UCTN)-

The conclusion of the CWG-UCTN on 2-character codes, and the rationale for that conclusion, was:

"The CWG recommends that the existing ICANN policy of reserving 2-letter codes for ccTLDs should be maintained, primarily on the basis of the reliance of this policy, consistent with RFC 1591, on a standard established and maintained independently of and external to ICANN and widely adopted in contexts outside of the DNS (ISO 3166-1)."

Full report here: https://ccnso.icann.org/sites/default/files/field-attached/ccwg-ctn-final-paper-15jun17-en.pdf

Kind regards,

Heather Forrest

Gnso-newgtld-wg-wt5 mailing list
Gnso-newgtld-wg-wt5 at icann.org<mailto:Gnso-newgtld-wg-wt5 at icann.org>

Gnso-newgtld-wg-wt5 mailing list
Gnso-newgtld-wg-wt5 at icann.org<mailto:Gnso-newgtld-wg-wt5 at icann.org>

Gnso-newgtld-wg-wt5 mailing list
Gnso-newgtld-wg-wt5 at icann.org<mailto:Gnso-newgtld-wg-wt5 at icann.org>

Gnso-newgtld-wg-wt5 mailing list
Gnso-newgtld-wg-wt5 at icann.org<mailto:Gnso-newgtld-wg-wt5 at icann.org>

Gnso-newgtld-wg-wt5 mailing list
Gnso-newgtld-wg-wt5 at icann.org<mailto:Gnso-newgtld-wg-wt5 at icann.org>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mm.icann.org/pipermail/gnso-newgtld-wg-wt5/attachments/20180119/8e84123f/attachment-0001.html>

More information about the Gnso-newgtld-wg-wt5 mailing list