[address-policy-wg] Registry - not a policy proposal

Piotr Strzyzewski wrote:

> On Tue, Jun 01, 2010 at 08:43:49AM -0700, Leo Vegoda wrote:
> 
>>>Ref. Section 5, "comprehensive":
>>>
>>>From a "consumer's" point of view, the RIPE NCC Address Registry should
>>>return a valid answer about the status of *all* addresses in the IPv4 space.
>>>
>>>The answer will necessarily be different, depending on the status of the
>>>address block and the authority regarding the answer.
>>>
>>>For those addresses, which are authoritatively managed by other RIRs, there
>>>should be an indication where to find "better" information. We are already
>>>pretty far down that path :-)
>>
>>The 'user friendly' thing to do is probably to go and find them the
>>end answer, rather than referring them to somewhere else to look up
>>the answer. This is probably particularly important when the query is
>>made by a piece of software that is not aware that it needs to follow
>>a referral.
> 
> 
> I believe that this is not always the best solution. Although this is
> generally good idea when user is querying registry using some API with
> well-defined output, I can imagine that simple redirection of text-based
> whois output from other registry can confuse the querying software.

Well, redirection is one possible appraoch, actually an approach I do not
really like ;-) Another possibility would be to have the targetted whois
server do the legwork on behalf of the user. Then "just" give back the
(correct(def.?)) answer.

I presume this would require a considerable effort on the side of the RIRs
(and NIRs, potentially using a different character set) and a much better
understanding ref. the (compatible) format (and the components) of the result.

> Piotr

$ set (dreaming mode=OFF),
Wilfried