From: Bruce Tonkin 
Sent: Friday, 8 June 2001 7:14 AM
To: 'jo.lim@auda.org.au'
Subject: Response to the New Second Level Domains Public Discussion
Paper

To:
Ms Jo Lim
Chief Policy Officer
.au Domain Administration


Melbourne IT response to the New Second Level Domains in the Australian DNS paper 
============================================================================
=====

(1) What criteria should we use in considering new 2LDs

Many of the existing 2LDs within ".au" have fallen into disuse (e.g conf.au and info.au). 
This has either been due to the lack of marketing by the registrar to inform consumers of 
the existence of the domain, or the restricted ability for an individual operating as a 
registrar to devote sufficient time to provide a
reasonable level of customer service for the domain name space.   The 2LDs
(e.g com.au and net.au) operated on a commercial basis presently have the greatest consumer 
acceptance.

Thus a criteria for considering a new 2LD should take into account the ability of auDA or 
the proposing individual/organisation to invest sufficient resources to adequately market 
the existence of the new 2LD, and provide a sufficient level of service (as discussed in 
the competition panel report).  Simply adding new 2LDs without associated marketing and 
technical resources, will unnecessarily clutter the ".au" space.

Melbourne IT recommends that auDA consider incorporating some of the criteria used by 
ICANN in selecting the new TLDs (http://www.icann.org/tlds/tld-criteria-15aug00.htm.  
These include:
* the extent to which a new 2LD would lead to an effective "proof of concept" concerning 
the introduction of additional 2LDs in future (e.g try "doctors.au" before introducing 
"lawyers.au")

* the enhancement of the utility of the DNS (e.g use of the DNS using geographically or 
numerically coded identifiers to point to physical Internet locations)

* the completeness of proposals and the extent to which they demonstrate realistic business, 
financial, technical and operational plans and sound analysis of market needs (note this 
does not exclude non-profit organisations for submitting proposals)

New TLDs may either be for general use (such as ".com.au") or have a specific use (e.g 
"doctors.au").  A proposal for specific use should ensure that the proposer is sufficiently 
representative of the expected user group (e.g the AMA rather than an individual with no 
association with the medical profession).

(2) How should existing 2LDs be changed?

Part of the problem with the existing domains has been identified above. The industry 
restructure described in the competition panel report, will assist in improving the service 
levels for the 2LDs.  

The Names Panel report mostly focussed on rules that apply to ".com.au", and attempted to derive 
from the ".com.au" rules a set of generic rules for ".au".  Melbourne IT recommends reconvening 
the Names Panel to review the policies in detail for domains that have fallen into disuse.  
These include "info.au" and ".id.au".  The introduction of a ".info" and a ".name" top level 
domain is likely to stimulate demand for these domains within Australia.  auDA can leverage the 
huge marketing budgets being used for the new TLDs to the advantage of 
the 2LDs in Australia.  Both ".info.au" and ".id.au" have had overly restricted policies.

Domains that have low levels of usage should be gradually phased out (e.g via not accepting new 
names).  For example ".conf.au" may not have a long term future, unless it is marketed to 
conference management organisations more effectively.  Perhaps a directory associated with 
".conf.au" with a program of conferences in Australia may make this domain more attractive.

Thus the process used by the Names Panel for ".au" can be continued to look at individual 2LDs.  
This is probably best done one at a time, with appropriate opportunity for public comment.


3.  What new 2LDs should be introduced?

Individuals - this should be a high priority as individuals are currently poorly served within 
".au".  Either align ".id.au" to the policy of ".name" for consistency, or create a new 2LD called 
".name.au".

Open slather 2LD - this should be a low priority.  This will lead to a flattening of the ".au" 
domain space and a break down of the utility of the DNs which is designed to support a hierarchy 
of domain spaces with specific purposes.  It is also likely to lead to extensive cybersquatting 
and consumer confusion as we currently see with ".com"

Geographic names 2LD - care should be taken not to create a hierarchy that is too deep, and 
hence unpopular with users - eg street.suburb.council.capitalcity.state.au. Note the experience 
of ".us".  

Gateways - should be introduced on a proof-of-concept basis initially, to avoid creating consumer 
confusion before the concept is proven.

New commercial domain - this should be a low priority.  Part of the reason for introducing 
".biz" was the perception amongst users that all the "good" ".com" names have been taken. This 
is not the case with ".com.au", where most organisations can get the domain name they desire 
subject to eligibility rules without a major concern from already registered domain names.  
".biz" is intended to be what ".com" should have been, ie a name space for businesses rather 
than an open slather domain with no purpose.  ".com.au" already provides the business space 
effectively in Australia.

The other suggestions are all equally viable and should be assessed as part of the process 
described in (1).


4) What should be our response to the new international TLDs?

The domain names that have the most usage in Australia are those that match the top level 
domains (".gov", ".edu", ".com", ".net").  There is a high degree of consumer awareness for 
these namespaces due to the large usage of these domains internationally.  ".au" should 
continue to follow the international top level domains where possible.  In particular ".au" 
can regenerate interest in ".info.au" by taking advantage of the large marketing campaign for 
"info" internationally.  ".au" can also regenerate interest in an individual name space - 
e.g ".name" or ".id".  Other new top level domains such as ".museum" and ".coop" could be 
introduced in Australia after the policies and demand for these domains is clearer.  The level 
of marketing of these domains is likely to be lower than for ".info" and ".name".

5) What process should be used to propose new 2LDs?

There should be a separate process to consider the current 2LDs - ie via the names panel.

Melbourne IT believes that a process where submissions are made for new 2LDs is appropriate. 
In contrast to the view of the panel, Melbourne IT recommends that the ICANN criteria for 
" the completeness of proposals and the extent to which they demonstrate realistic business, 
financial, technical and operational plans and sound analysis of market needs" is an important 
one.  The panel needs to recognise that the success of a new 2LD depends as much on the marketing, 
technical and operational plans as it does on an "analysis of market needs".  Past history in 
".au" and other cctlds has shown that 2LDs will fall into disuse unless there is a sound model 
for their development.  Note that this does not exclude non-profit or public institutions (e.g 
Australia Post, government departments) from submitting proposals.

In a competitive registry environment, a sponsoring organisation could choose to partner with 
a particular registry operator.  In a single registry environment it would be important to get 
the support of the registry operator with an understanding of the costs of running the new 2LD.

A proposal should consider how much to charge for a domain name (it may be free), and this 
should reflect the costs in providing the service described in the proposal.  This is particularly 
important in policy rich domains, such as ".com.au" where the business model needs to scale.  
For example a new domain name space can be provided by an individual for free provided the 
volumes of names registered is low.  Once the volumes increase the service levels rapidly drop 
off. A proposal should consider the expected volume of registrations and a model for how the 
policy for the domain will be carried out.

Domain name proposals without a sound model for operation are destined to fail.  Incorporating 
the ICANN criteria will make the process of managing a huge number of proposals more reasonable, 
and ensure that proposals are well thought out.  auDA could elect to use a lower application fee, 
and take advantage of voluntary panels to evaluate the applications.  The 47 proposals for the 
ICANN top level domains are publicly available on the ICANN website and will serve as a useful 
starting point for potential applicants. An organisation that proposes a new domain name space 
should be prepared to take on part of the risk for developing a new 2LD (ie put up marketing 
funds to raise consumer awareness, and invest in sound market surveys), to avoid undue risk 
being placed on the rest of the domain name industry (e.g auda, registries, registrars, and 
resellers).  It is reasonable to expect an organisation to be able to achieve a reasonable 
return on their investment to compensate for the risks undertaken. Note again that this does 
not exclude non-profits or free services, it is only inclusive of for-profit.

Dr Bruce Tonkin
Melbourne IT Ltd