[gnso-ff-pdp-may08] A question related to proposed solutions

[RLVaughn <RL_Vaughn@xxxxxxxxxx>]

Mike O'Connor wrote:
> Seems as though we've got a pretty good handle on the definition.
>
> How about opening up the floodgates on a discussion of proposed solutions?
>
> We have a few proposals.  Are there others?  I know that I've sortof 
> damped this part of the conversation down a bit, while we were hammering 
> out the definition.  Now's the time to present alternatives.
> m
I would suggest we also spend just a bit of time asking ourselves about
identifying and verifying problematic domains at the same time we
are addressing solutions.

Corrective actions depend not only on what we know but also when we
know it.  Earlier, accurate, discovery can lead to more proactive
solutions than reactive domain shutdowns.

From my Storm example, I would assume the gtld server operators would be
in a very good position to detect highly volatile name server assignments.
Is that a valid assumption? If so, would it be possible to instrument the
gtld servers in order to take advantage of the visibility?  If such 
instrumentation is technically and economically feasible would it be possible
to rapidly contain undesirable behaviors or to provide some background
feed to registries about questionable domains?

Failing that, would it be in-scope to consider having the gtld servers
honor their own advertised TTL for name servers?  Would it be in-scope
to have a gtld-based fee structure for those domains that want to
change gtld records frequently?  If those are even meaningful solutions
what incremental harms do they cause?