A Lurker's comments

[Peter Murray <pem@po.cwru.edu>]

Executive summary:
  * I have no intention of running or being involved with the operation
    of a new registrar or a steward (as defined by the Dec19 Proposal).
  * In favor of the shared registry scheme.
  * In favor of the new randomly chosen order for bringing new 
    registrars online as proposed by Kent Crispin.
  * In favor of the CORE group choosing new gTLDs.
  * Opposed to 60-day waiting period between domain registration and
    activation in the DNS.
  * Support the approximate US$20,000 fee.


So that you know where I'm coming from...I am not/will not be involved in the
operations of a registrar, but being in the position that many of you find
yourself in, I have to explain the workings of the Internet to staff
(librarians, in this case), customers, and my mother (not an easy task).
Therefore, my prime concerns are with the stability and consistency of 
network functions.


On to the issues...First, I believe the proposed shared registry scheme is
the best for the health of the Internet.  I've thought about this long and
hard, and it comes down to two issues.  On the one hand, if the gTLD was
under the sole control of one organization, that organization would want to
do the best job it could to promote value in cost effectiveness,
responsiveness, and stability in order to differentiate itself from other
gTLD and registry operations.  And given my prime concern about appearance to
the Internet users, the registrar organization has a lot of control over how
stable their DNS zone will appear because they have all of the decisions
about locations of servers, number of secondary machines, etc.

On the other hand, in a shared registry scheme much of the concern about how
the zones appear to the Internet users are driven by the actions of the CORE
group and the steward for the gTLD.  A registrar, while concerned about this,
cannot directly influence it.  Since the organizations who have a direct
financial interest in how the zones are handled can't directly control their
operation, problems are likely to take longer to resolve (eg, any government
bureaucracy).  But the advantage of the shared registry is that the SLD
holder can move their registration between registrars of a gTLD when one
registrar fails to live up to their commitments.  Since a SLD holder will
have obviously put a lot of work into promoting their domain, this is a GOOD
THING.  Competition is promoted between registrars of a gTLD to provide the
best price/performance possible (also leaving open the opportunity for
different levels of prices for different levels of performance or bundled
features, such as web homing, etc.) 


Secondly, there have been a lot of complaints about the lottery process for
choosing registrars for the new gTLDs.  I think that Kent Crispin's proposal
for a random ordering of acceptable candidates is a good one, and I've
noticed that the discussion on this topic has died down since his posting.
This is also a GOOD THING because it implies consensus in the list
membership.  If you miss his proposal, it can be found at:

   http://www.iahc.org/iahc-discuss/mail-archive/2631.html


Thirdly, I'm in favor of the CORE coming up with a list of new gTLDs that
will become part of the official root zone.  Once again, in terms of
consistency to users, this function needs to be driven from a global
perspective, and not that of what will make money for a registrary.  If gTLDs
are not consistent and predictable to users, the whole Internet becomes
harder to use.  The phase in of new gTLDs (such as the number '7' proposed in
the draft) over time also promotes this premise because it will allow users
of the network to gradually adjust to new domains rather than be swamped with
a hundred of them all at once. 


Fourth, I'm opposed to the 60-day waiting period, but mainly for reasons
discussed on the list.  It does not seem to have legal weight.  It makes the
whole registration process cumbersome.  And I'm not convinced it is needed. 


Lastly, I support the US$20,000 (or approximate) fee.  I support this
primarily because CORE is going to need funding to hire stewards who will do
a good job in putting the zones on the network, and US$500,000 sounds like a
good starting point to do this (25 registrars at US$20,000/each).  In fact,
it will probably take several years before the CORE breaks even in order to
build the infrastructure to move forward.  As a side effect, US$20,000 is a
sufficient but not insurmountable entry barrier along with the other
technical requirements outlined elsewhere (network connectivity, etc.). 


That's about it from this lurker's point of view.  I'd be happy to go into
more depth about any one of these points. 

Happy New Year!

Peter
--
Peter Murray, Library Systems Manager                      pem@po.cwru.edu
Library Information Technologies         http://www.cwru.edu/home/pem.html
Case Western Reserve University, Cleveland, Ohio            W:216-368-8989