Re: The view from my window

[davidk@ISI.EDU]

Hi Dave and Hank,

> Dave Crocker writes :
> 
> At 6:20 PM -0800 12/22/96, Aveek Datta wrote:
> >I also think the lottery is a bad idea, almost as bad as the FCFS which
> >all these experimental registries advocate. The best is to draft some
> >sort of ranking system (0-100 Point System?) for registries and then
> >accept the highest ranking registry. On ties within a certain fuzzy
> 
> 	Either the ranking system is highly -- i.e., completely --
> objective and therefore won't receive much criticism, or the scheme will be
> viewed as highly subjective.  My own guess is that highly objective
> criteria will be too few to adequately partition applications.  I think we
> have no chance of using strictly objective criteria.  Those believing
> otherwise are invited to devise and submit a scheme and have it discussed
> on this list.
> 
> 	Given this, an additional mechanism is required.  The only way to
> choose from among an equally qualfied set (according to objective criteria)
> is a random process.

I fully agree with this.

I would even go one step further:

We should have minimal requirements to give as many people as possible
the possibility to enter this business. Making too detailed and high
requirements will hurt more then it solves. This allows for people to
provide cheap/expensive and crappy/good service. Also, as I pointed out
early, a registar might abide to the rules at day one but a few months
later loose some key personnel and not be qualified anymore. Or somebody
hires a consultant to get through the qualification and then take over
themselves. It will then turn out that it is much more difficult to
revoke a license then to give it out.

Hank challenged me to come up with some requirements that are simple and
objective. Let's give it a try now:

- Fixed application fee (whether it's 20K or another amount as long as it
  covers the costs and a bit more to give some security level for the
  whole system CORE/IAHC)

- IAHC makes a test gTLD (or test SLD) domain.  IAHC gives the applicant
  access to the (test) repository IAHC applies for three test domains
  with the applicant and see if they get things installed correctly from
  a technical point of view within a month.

Note that it is important to know whether an registrar will also runs
secondary for the gTLDs (not covered in draft). Technical competence is
much more important then.

These kind of criteria can be verified and are not subjective.

I invite everybody to come up with better ideas as long as they can
easily verified and are objective (this will also keep the application
fees low).
  
David K.
---