RE: [gnso-irtp-b-jun09] i'm offering another approach to the ERTP thingy

["Diaz, Paul" <pdiaz@xxxxxxxxxxxxxxxxxxxx>]

Can you please attach SAC007 (I don't remember what the Emergency Action
Channel proposal entails)

 

 

________________________________

From: owner-gnso-irtp-b-jun09@xxxxxxxxx
[mailto:owner-gnso-irtp-b-jun09@xxxxxxxxx] On Behalf Of Mike O'Connor
Sent: Friday, February 11, 2011 4:16 PM
To: <Gnso-irtp-b-jun09@xxxxxxxxx> List
Subject: [gnso-irtp-b-jun09] i'm offering another approach to the ERTP
thingy

 

hi all,

 

i've shamelessly edited the document we reviewed last week, making it
even simpler but i think also better.  i'd like to take a few minutes on
the call to review this idea.  here's the gist of it.

 

the problem with ETRP and all of it's related kin is that it:

 

        a) forces the registrar and/or the registry to judge the merits
of a hijacking claim at lightening speed -- and essentially makes them
dispute-resolvers in a situation of imperfect information.

         

        b) that high-speed judgement also leaves the process open to
gaming by disgruntled sellers looking to "claw back" a domain name.

 

preventing these two problems leads to all kinds of complexity and, as
we discovered, eventually gets too heavy for its own weight.  so let's
step back...

 

what if we:

 

        a) recommend enacting the Emergency Action Channel proposed in
SAC007

         

        b) remove the judgement-call/dispute-resolution by
registrars/registries

         

        c) make the criteria for urgent-return be factual and
externally-verifiable -- "the gaining registrar does not respond to an
Emergency Action request within X hours"

         

        d) remove the changes to the TDRP, since the return is being
done for non-disputable reasons

 

this way, neither the losing registrar nor the registry have to evaluate
anything to do with the claim of hijacking, they simply are provided a
way to get a domain name returned to its prior state in the case where
the gaining registrar is not communicating.  presumably, since
registrars are pretty confident that they can resolve most of the
hijacking problems if they can actually *talk* to the other party, this
would address some pretty large proportion of the problem.

 

what say you?  i've attached the slashed up version of the draft for you
to review.

 

mikey