[Ietf-not43] -02 c's and q's...

[Ted Hardie]

> 
> ...in no certain order....
> 
> 3.2.3 - I'd like to recommend that we change the requirement to read: "The
> service MUST provide a _lightweight_ mechanism to distribute this search
> across applicable domain registries and registrars. The service SHOULD have
> a means to narrow the scope of a search to a specific TLD.  _The service
> MUST provide a mechanism to restrict the scope of a search to specific TLDs,
> registrars and registries. The service SHOULD provide a mechanism for
> operators to opt-out of this search or series of searches. The service
> SHOULD NOT distribute queries to operators that have opted out. The service
> MUST be able to specify to the client an empty result set should the search
> yield no results."
> 
> Comments: I think the lightweight language speaks for itself, hopefully it
> addresses some of Rick's concerns (with which I generally concur). As to the
> rest, there are any number of operational circumstances where I don't want
> to answer queries for a definite or indefinite period of time  - if I don't
> want to answer, it seems rather heavy to keep pushing requests my way even
> if I am the appropriate place to find the data. I do have a further concern
> that there should be an opt-out mechanism for registrants to opt-out of
> specific or all searches based on their individual preferences. I need to do
> a little more thinking and reading on this one though (but figure that I
> shouldn't hold on to the general concept privately any longer ;)


I personal like the "lightweight" language above, but I am concerned
about the opt-out mechanism.  If the mechanism is simple enough to
opt out without requiring a server-to-server interchange, I think
life will be better; I also don't think we can use crisp as the
protocol to conduct that interchange, so I am leary of the requirement.

				regards,
					Ted Hardie