Second Level Domains Policy Review Submission

From:		Ewen McNeill
Received:	6 October 2003


There are at least three different feedback mechanisms suggested for commenting 
on this proposal, and it is by no means clear which of them are actually being 
taken into account and which ones are simply there to keep people amused and 
actually go to /dev/null.  

So I've chosen two: the very underused nz-review mailing list (I had thought 
that the working group was going to use it, but either they have not used 
email at all, or it has been deliberately kept out of the mailing list), 
and policies@dnc.org.nz to ensure that it is not overlooked.

I'll respond to the questions raised in the discussion paper at:

http://dnc.org.nz/content/2ld_background.pdf

as a way to focus my comments.  (Again I express my concern at the use of 
PDF for an almost entirely textual document and the difficulty it creates 
in quoting text to respond to.  Any typos not in the original document are 
due to being forced to retype all the questions while swapping between 
windows.)


Issue 1: Existing Second Level Domains

1.  Are the twelve second level domains still appropriate for their 
communities of interest?

With the benefit of hindsight both iwi.nz and cri.nz might not have 
been created (with maori.nz and co.nz possibly being more appropriate 
homes).

However while they still have registrants they should be retained.  If 
any 2LD should ever drop to 0 registrations, or all remaining registrants 
agree they would prefer to transfer to another name, then consideration 
should be given to removing the (unused) 2LD.

school.nz also might not have been created, but the rationale for 
distinguishing between school.nz and ac.nz was basically sound.

FWIW, renaming 2LDs for the sake of renaming them should be strongly 
resisted.


2.  Should any of the open 2LDs be moderated?  (.ac, .co, .gen, .geek, 
.maori, .net, .org, .school)

ac.nz and school.nz might benefit from moderation if suitable rules 
and a moderator could be found (they should, however, be fairly 
inclusive and not just limited to, eg, "public schools run by the 
Government").

co.nz is too much of a mess (due to the lack of other 2LDs being added 
at appropriate times) to even consider moderation.  (A ltd.nz style 2LD 
consisting only of the registered company names of registered companies 
might make sense.)

org.nz is also too much of a mess to even consider moderation (a inc.nz 
style 2LD consisting only of incorporated societies might make sense).

net.nz is also too much of a mess to consider moderation, and the 
original rationale (naming network infrastructure) has largely been 
forgotten (even InternetNZ "abuses" it).

gen.nz, and geek.nz are inheriently "and the rest" open areas.  
maori.nz requested to be unmoderated and I believe that request 
should be honoured (although had it been me asking I might well 
have strongly considered making it moderated).


3.  Is there any need for rationalisation of the current 2LDs?

Yes.  Should it be done: no.  Not unless any of them fall into 
disuse.

2LDs that fall into disuse should arguably be removed.   2LDs 
where all registrants petition to have it removed should arguably 
be removed.

Why?  "Rationalisation" would be very disruptive and achieve only 
a level of "intellectual purity" which would soon be lost again.


Issue 2: 2LD creation process

4.  What criteria (types of criteria and specific examples) should 
be applied to the creation of new 2LDs?

New 2LDs should be created any time there is a demonstrated 
community of interest in the creation of the 2LD, and there isn't 
another 2LD that is (or would be if created) more appropriate.  
(If those representing the community of interest cannot easily be 
convinced that the alternative 2LD is more appropriate then it 
probably isn't more appropriate.) "Misc" 2LDs (such as gen.nz) 
should be excluded from consideration as "more appropriate".

The process should be considerably less time consuming, and less 
bureaucratic than the existing process.  A Usenet newsgroup 
creation style discussion approach is probably a reasonable model 
-- the proposed group is discussed for a while, some people try to 
convince the proposers that it'd be better with a more sensible 
name, etc, but after a relatively short space of time a decision 
is made.


5.  Is the current process for creating 2LDs too simple, too 
complex, or about right?

It is far too complex.  While the process should obviously not 
be a "1 day" thing, a 1 month process should be considered on 
the long side of desirable measured from the submission of a 
detailed proposal.



6.  How much, if anything, should an application for the creation 
of a new 2LD cost?

$250 (or maybe $500).  That, too, is a number plucked out of the air, 
but represents on the one side a desire to discourage "whims" that 
no one will support, and on the other side to encourage many more 
2LDs. The existing application cost, of $1000, acts as too strong 
a discouragement to new 2LDs, leading to poor management of the 
namespace as people invent many variations on names instead of 
creating an appropriate 2LD (eg, variations on "bank", when "bank.nz" 
would have been suitable).


7.  Who, if anyone, should be able to block the creation of new 
2LDs?  If so, why and how?

"Trademark holders" should be able to block the creation of 2LDs 
in their name against their will.  "Trademark holders" should be 
interpreted loosely so that, eg, individuals should be able to 
block the creation of 2LDs in their name against their will.  
(Arguably this should apply in any moderated domain space.)

Off hand I cannot think of anyone else who should be able to 
veto the creation of a new 2LD.


8.  Should there be different criteria or processes for the 
creation of moderated as opposed to open Second Level Domains?

A clear set of moderation criteria should be considered to be 
strong evidence of a community of interest for the creation of 
a 2LD.  Without moderation criteria this interest needs to be 
demonstrated in other ways (eg, with a number of potential 
registrants pledging their interest).

In addition the moderation criteria should be subject to the 
same discussion as the 2LD name itself (to explore whether 
slight different moderation criteria would make the 2LD 
significantly more useful; or different moderation criteria 
of an existing 2LD would make that more appropriate).

Otherwise the process should be similar for both.


9.  Is the level of public input in the process adequate; 
excessive; or inadequate?

The level of input is adequate.  The time allowed for it is 
dramatically excessive.  At very least there should be a provision 
by which an absence of discussion for some days allows progress 
to the next stage to be expedited so that once any discussion 
naturally draws to a close progress can be made without everyone 
forgetting about it (and starting arguing again 2 months later).

A 9 month process is far far too drawn out; as mentioned above a 
target of a 1 month process is considerably more appropriate.


10. Are the methods currently used to seek public input (submissions, 
discussion period, straw poll) the best available, or would you 
suggest other/additional methods?

The methods are fine; it is the time for which the process stalls 
at each of these stages even after the public has lost interest 
which is the problem.



Issue 3: Direct registrations at the Second Level

11. Do you think public registrations should be accepted at the 
second level?  Why?  What are the advantages or disadvantages of 
such an approach?

No, direct public registrations should not be accepted at the second 
level.  While it would merely be unwise to start a new CCLTD with 
direct registrations at the top level, it would be foolish to attempt 
to convert a CCTLD with an established hierachy of 2LDs into a "flat" 
namespace.  

If direct public registrations were available at the 2LD level, then 
it would be impossible to ever have a sensible hierachy of 2LD/3LD 
names again; the flat namespace would be permanent.

The mess in .com, .co.nz, etc, and particularly the many many 
variations on creating alternative names (eg "FOOthemovie", 
"FOOmovie", "movieFOO", etc), demonstrate the problems with 
attempting to force a non flat structure (human endeavour), into a 
flat namespace (*.nz).

There should definitely be many more 2LDs; but that should be to 
create more hierachies for "sensible" organisation, rather than 
to flood another namespace to the point where sensible allocation 
is no longer possible. (I would suggest that at very least there 
should be one 2LD for each trademark "category", plus several more 
for good measure.)

You don't ask -- but should have -- whether 4LD hierachy should 
be encouraged: it should.  A hierachy like wellington.rugby.sport.nz 
would be nice; as would FOO.wellington.reg.nz.


12. If direct registrations are allowed at the second level, 
what kinds of transition arrangements would need to be put 
into place for existing third level registrants?

Any holder of a 3LD should be given "first refusal" on any 
appropriate (ie, identical, confusingly similar) 2LD name; in many 
cases this might just be the same as the 3LD component of their 
2LD name but not always (eg, where there's already a conflict at 
the 3LD level due to lack of suitable 2LDs).  Some form of 
"pre-registration" process is probably the best way to determine 
self-identification with "appropriate" names, with a check that 
they are "similar" to the existing 3LD name.

Where two or more 3LD registrants want the same "similar" 2LD name:
(a) they should be encouraged to work together to reach an suitable
    arrangement (money may change hands, or alternative names be
    achieved, or some "sharing" process be adopted);

(b) if an agreement cannot be reached then _none_ of them should be 
    able to have that 2LD name _and_ no one else can have that 2LD name
    (except through the process described in (a)).

The resulting mess of the collapsed space would be irredemably messy 
anyway so there's little point in worrying about other transition 
arrangements to, eg, try to preserve sensible names or prevent a 
"land grab".  (Besides the "land grab" has largely occured in co.nz 
anyway, so would be subject to the first refusal approach above.)

I must say I sincerely hope that this doesn't have to happen; it'll 
be expensive, time consuming, and only achieve removing three 
characters from domain names (on average) at the expense of never 
being able to have a sensible hierachy.


13.  What is the best way to deal with any conflicts between 
existing third level registrants if the second level is opened?

As described above: force the registrants to come up with a solution 
they can all live with, and if they cannot, do not let anyone have 
that name. It is the only approach likely to produce solutions that 
everyone can live with.


14.  Should registration fees be different (higher or lower) for 
Second Level domain registrations if they are opened directly to 
the public?

They might as well be the same.  It won't make much difference, 
as third level domains will rapidly become "third class" citizens.


15.  Are there other options around the future of the Second Level 
that the working group should consider?

It should be treated -- as it has been -- as a moderated domain.  
It may make sense to consider allowing sensible 2LD-only entries 
where there is clearly no suitable hierachy underneath.

The only one I can think of at present is "www.nz", but there may 
be others in future.  (Who runs these things shared resources 
should be considered carefully to ensure they are done in a manner 
which would benefit the whole New Zealand Internet community.)

A wildcard *.nz entry should definitely NOT be put in place.

Aside from that the approach to enabling more 2LDs should be to 
make the application process more streamlined, not to abandon 
the common sense approach of building a sensible hierachy.


General Feedback

16.  Are there any other issues you think the Working Group needs 
to consider in the course of the review?

As I allude to above, the creation of structured 4LD names should 
be considered (eg, for regions, sports, people, etc), or at very 
least the mechanism by which this might be proposed.

It would also be worth considering proposing some moderated 2LDs 
like ltd.nz, inc.nz (both mentioned above) in a proactive fashion 
to suitable organisations that would be able to moderate them 
directly (eg, the Companies Office).  A regional hierachy (foo.reg.nz), 
and a personal hierachy (foo.id.nz?) may also be worth proposing 
proactively, particularly as well structured 4LD space.

Finally please keep in mind that you can't make the "complication" 
of "which 2LD" go away simply by doing away with 2LDs.  Instead of 
the "which 2LD" complication, you'll have the "what suffix or 
prefix did they use to make their name unique" problem.  

Trading an easier problem ("which 2LD of the limited set") for a 
harder problem ("how inventive could they have been in coming up 
with a unique name") is rarely a good idea.  Namespace collissions 
will remain irrespective of what you do, and IMHO hierachy is the 
most sensible way to organise things to reduce the effects.  
Trademark law worked this out over 100 years ago; those who do not 
remember history are doomed to repeat it.


Ewen McNeill