<<<
Chronological Index
>>> <<<
Thread Index
>>>
Re: [soac-newgtldapsup-wg] Draft summary of public comment (second milestone report) ready for JAS WG review
- To: michele@xxxxxxxxxxxxx
- Subject: Re: [soac-newgtldapsup-wg] Draft summary of public comment (second milestone report) ready for JAS WG review
- From: ebw@xxxxxxxxxxxxxxxxxxxx
- Date: Wed, 03 Aug 2011 13:25:42 -0400
Michele,
Thanks for the forward from your co-worker. I see we share the same old kit,
archaic cisco pix boxen.
What he misses isn't the ease of deployment, by some tunnel provider, it is
the assumption that "decent transit provider" are ubiquitous, that if pushed
out of an airplane over the Matto Grasso or Burundi you could find one in
finite time.
The discussion has never been about rational means to solve a problem in
commercially reasonable constraints, but an absolute drop dead requirement
at the transition to delegation point, which has only a year extension at
the applicant's request.
Now if we assume that the applicant is using some implementation we already
know is operating on a v6 network, the only cost that Bill (and others who
make a "harm to applicant" claim) can point to is the cost to sixify the
local kit, about an hour's time for your guy, and more if it is remote hands
and the configuration problem is done with a "sixify asset" over the phone.
So now the stark problem is Bill's claimed that the applicant will be harmed
if allowed to turn up the service, and we know the cost of not turning up
the service is about $200,000 USD, and that's so parsimonious on the non-fee
costs as to be untenable or assume a shrink-wrap source for the application.
Our best guess at the cost of sixification, when a "decent transit provider"
arrives at the applicant's venue, is an hour of competent time. There is a
significant disconnect between the harm known, and the harm claimed.
One can reasonably claim that v6 availablity makes the available infrastructure
reliable, whether v6 is used or not, just as one could claim that fiber makes
the network reliable, or some degree of redundency, or ... and therefore the
siting of a registry is disallowed by policy until which ever of these is
selected as the predicate condition is actually realized.
I don't for a moment think that you've changed your view, I do want to ask
though if because "v6 is easy" (to paraphrase your prior combined note),
it remains a reason to decline to recommend to the Board and the eventual
policy implementors that if an applicant is located in a venue where v6 is
not available, that the v6 requirement be deferred?
Eric
<<<
Chronological Index
>>> <<<
Thread Index
>>>
|