From:           Ewen McNeill
Received:     1 June 2012

Question 1. Should the New Zealand domain name space be extended to allow registration at the second level, for example
This suggestion has been raised repeatedly in .nz (most recently that I recall a few years ago), and each time the conclusion has been that it would create a bunch of problems (some of which the consultation paper highlights) for little benefit. Which still seems to be the case. The cost/benefit ratio of making this change seems very poor, particularly when all the externalities (ie, effects on those not benefiting) are taken into account.
Discussion on the NZNOG mailing list managed to identify two benefits:
1. Some people would get to leave, eg, ".co" out of their domain name; 2. NZRS (and by extension InternetNZ) would get a windfall profit as a result of anyone who cares about their trademarks needing to register 1-3 additional variations on their existing domain names (eg,,, etc).
By contrast the cost is not only additional registration fees, but increased risk of someone else having "look alike" sites (eg, phishing), changes in "business stationery", educating the public, an entire process to deal with taking in new registrations of potentially conflicting names, etc.
If this was being considered in the early days of wide spread adoption of the Internet (eg, mid 1990s) then perhaps a case could have been made that the benefit outweighed the cost of changing. Without really compelling benefits from this change (and "NZRS will double its revenue" does not strike me as a compelling reason) this really is not the case now.
In addition making such a change abandons the ability to have a "curated" 2nd level domain structure for .nz (as has been done for the past 20 years); this is a "flag day" change and it would be impossible to back.

Question 2. Are there any other undertakings that the Domain Name Commission should make while developing/implementing the policy?
If the DNC is determined to do this, then it should create policy to reduce the harm caused by its decision including, eg: 1. Forbidding any domain ending with "-${2LD}" (for all existing 2LDs of .nz), to avoid each registration having to get "typo" variations of their name like "".
2. Ideally extend to the above to all common typo variations of those 2LDs (eg, "") to reduce domain squatting harm
3. Prohibit registration of various automatic service discovery names (eg,; there are various others)
4. Block (or turn into a service providing advice) common service names (eg,,, etc), particularly to deal with DNS resolvers that walk up the DNS hierarchy (IIRC I've already been caught out by like this, and had to implement DNS lookup workarounds for it; it would be worse with open registration in .nz)
5. Ideally make the cost of registering a 2LD (rather than 3LD) many orders of magnitude higher (eg, root top level domains are going to be $100,000+ versus about $20 for one level down; IMHO the proposal that 2LDs in .nz be the same cost as 3LDs will make the domain speculation much worse than it would be if the speculators were required to invest substantial amounts of money in the process up front -- but then .nz's payment policies seem practically designed to encourage rampant domain speculation -- particularly the "return it 'hardly used' after a few days and it doesn't cost you anything" style of "domain tasting".)
6. Ensure all software vendors (eg, operating system vendors) are given sufficient notice to change their security assumptions around .nz subdomains from "three parts required" to "two parts required, except in this special list of 2LDs". I'd estimate they'd need 6-12 months minimum, perhaps 2 years.
7. Prioritise educating the public, and existing domain holders, about the change, the timeline for the change, and the process by which they can reduce the harm caused by the change to themselves and their organisations.
8. Cause NZRS/InternetNZ to widely publish a document detailing how much money they made as a result of the created "gold rush", and how that money has been used (and ideally the cost of implementing the change process, to illustrate how much money was wasted by this change).

Question 3. Should new second level domains be created to cater for particular interest groups, such as or
Absolutely. There is clearly a need for many more 2LDs in a well curated domain hierarchy. Up until InternetNZ took over .nz, new 2LDs were created pretty much on an "as needed" basis by people with reasonably good taste in namespaces. After InternetNZ took over the process became so bureaucratic and extended that the few 2LDs that did get created are a testament to persistence of those proposing them. It's difficult to measure how many didn't even get proposed because of the 6+ month process involved, with no certainty of outcome regardless of merit.
(, given as an example in the consultation paper, is sadly a lost cause, and an example of the harm caused by the excessively bureaucratic process: it took three proposals, years apart, to get created, and when it finally was the world had moved on so much that the intended users felt commercially unable to change from their existing domains.)
Given that is the largest 2LD in .nz, it seems obvious to me that there should be many more 2LDs that are subcategories of that, perhaps by industry grouping. Trademark law (which has dealt with these naming issues for centuries) might provide a useful guide to meaningful industry grouping.

Question 4. Should new moderated second level domains be created to cater for domain names that require special protection, such as
Where there is a clearly defined group of registrants _and_ the public interest would benefit from being able to clearly identify those users _and_ the group of registrants is commercially, politically, logistically, able to take up usage of the newly created 2LD (and commits to doing so), then absolutely yes such a 2LD should be created.
As noted above, is now a lost cause; commercially the intended registrants are too tied to their existing domains (and concern about phishing, etc) to be willing/able to change. is a good example of a well managed 2LD, that I would support creating again now. There are undoubtedly others were there is a clear test for whether something is or is not to be included/allowed in the set of valid registrants, where a moderated domain would make sense.

Question 5. Should the registration of some names such as or, be prohibited at the second level to minimise potential confusion? What names, if any, should be prohibited?
See also the earlier list regarding policies, some of which are repeated here:
... (many other variations of existing 2LDs)
Plus typo versions of all existing 2LDs. Plus anything that might look like intended hierarchy (eg,,
Plus all service domains (, and the like) used for automatic discovery by various services (AFAIK there's no central list of these, but from time to time new names of this type are found typically as a result of security issues they cause when registered at the "wrong" level).
Other common service names, eg "", "", "", and anything else that a user might typically give valuable passwords to.

Question 6. Do you agree with the rationale for the Sunrise Period that would enable existing .nz domain name holders first chance to register names at the second level? Why?
In general I agree that if such a change in naming policy were to take place at such a late stage (ie, 15 years into mainstream adoption!), then there needs to be a clear procedure by which organisations and individuals can claim the "same" name in the new naming conventions.
Where there is only a single registrant of FOO.*.nz (ie, for any 2LD) then it clearly makes sense that they have first right of refusal on that name as a 2LD.
Where there are multiple registrants of the "same" name in different 2LDs, the situation is less clear (and this is essentially another form of harm created by this policy change that is collapsing many different namespaces, potentially for different things, into one; Trademark law has never attempted to do this, for instance). I think in this situation the first attempt should be to try to reach consensus amongst the various registrants, which might possibly include them agreeing to "auction" the name off amongst themselves (eg, closed tender, with the only viable bidders those with an existing interest in the name already); that could plausibly happen without the DNC's intervention, but as creator of the problem it would be smoother if the DNC were also part of the solution.
In general I disagree with the idea (mooted on the NZNOG mailing list) of using registration date as a priority measure: for one thing, first registration dates were poorly recorded/transferred into the DRS, and for another the "different areas of endeavour" nature of different 2LDs means that in practice the first registration of a more generic name (eg, personal name/surname, generic service, etc) may not be any more legitimate than a later one.

Question 7. Who should be allowed to register a domain name at the second level when there are competing registrations at the third level?
I think the existing registrants of the same 3LD in different 2LDs should be entitled to a different "competing Sunset registration" process, to determine the appropriate (if any) registrant entitled to priority of the name as a 2LD.
I definitely do not think that these competing 3LDs should just be thrown into a First Come, First Served system. Such systems work very poorly in the case of a "open the gates, start now" gold rush, as has been demonstrated time and again by various systems (eg, ticket sales systems). It leads to unsustainable contention.

Question 8. Assuming only persons with a conflicting third level domain name may apply, how should that conflict be resolved? By consent? Or some other mechanism?
Other (below)
I think the initial process should be electronic mediation amongst the existing registrants, to have them agree on either:
(a) one of them that is entitled to priority;
(b) a person/organisation that they agree to be bound by their decision as to priority (eg, JP, lawyer, professional mediator, etc)
(c) a process by which priority will be determined (eg, first registration, most name recognition, etc)
(d) a process by which the conflict will be resolved (eg, closed tender for the name)
(e) that the registration be blocked forever until all conflicting registrants come to some other agreement
Absent that agreement, I think blocking it forever until all conflicting registrants come to an agreement should be the default option.

Question 9. Should the Domain Name Commission consider extending its Dispute Resolution Service for a limited period to cover particular sub-domains when considering whether a name registered at the second level infringes a complainant’s rights?
Clearly there is going to be more potential for "sound alike", "look alike" etc names created by this policy. I believe if the DNC is going to create this mess then they should use some of the windfall funds created by the change in policy to help clean up some of the created mess, rather than just leaving it to the courts/"market forces".

Question 10. Is the approach as outlined in the proposed amended policy in Appendix C appropriate? Why?
It's a good start, particularly if "appearance of being registered at the 3LD" is interpreted widely. I suspect additional policy review would be required if a change of this magnitude were adopted.

Question 11. Are there any other comments you would like to make relating to this consultation?
It is very late in the piece (about 15 years too late, by my count) to be making a change of this magnitude. It will cause considerable confusion, and involve considerable expense both to implement the change and deal with the external effects of having implemented it. To my mind there would need to be a truly compelling reason to create such disruption. There should be a much clearer description of the problem that is being solved by this change before it is given any further thought. "Some people want .nz 2LDs", and "we could make lots of extra money" are not particularly compelling reasons to me.
I would instead urge the DNC, as I did at the last review of such a proposal, to make it easier to create additional 2LDs that create sensible hierarchy. A well curated 2LD hierarchy with 50+ well chosen 2LDs, each with a few tens/hundreds/thousands of 3LDs under them would be much preferable to throwing thousands of conflicting names into one pool and expecting people to deal with conflicts by "adding more words on". One need only look at, eg, movie websites as example of the mess created: was it "", "", "", "", or something else. With a clear 2LD for, eg, movies, the answer would be much more obvious.