To:
"Hollenbeck, Scott" <shollenbeck@verisign.com>
cc:
"'Peter Eisenhauer'" <eisenhauer@schlund.de>, ietf-provreg@cafax.se, brunner@nic-naa.net
From:
Eric Brunner-Williams in Portland Maine <brunner@nic-naa.net>
Date:
Tue, 10 Apr 2001 08:20:07 -0400
In-Reply-To:
Your message of "Tue, 10 Apr 2001 06:54:59 EDT." <DF737E620579D411A8E400D0B77E671D75090F@regdom-ex01.prod.netsol.com>
Sender:
owner-ietf-provreg@cafax.se
Subject:
Re: 3.4.9 [4]
> This particular wording was requested by Eric Brunner-Williams. Eric, would > you care to clarify, please? Sure. Here is the excerpt from my original (off-list) note to the author of -06 (now -01) on February 8th: : 3.11 Object Information Query : : The return values of an information query (FQDN, NS, RegId) is detailed : and the query mechanism is registrar indifferent, except that the most : recent authorization identifier return is registrar specific, and any : extension of registry specific semantics is on an opt-in basis by the : incumbant registrar. : : Please invert the optional registry specific semantic extension so that : opt-out is the default, e.g., s/limited to/granted by/. : : [4] Requests to retrieve information describing a registered object : MAY be _granted_by_ the registrar that currently sponsors the registered : object. In case memories aren't perfect (and mine usually isn't, at least about other people's issues), the -06 draft had default opt-in disclosure semantics on queries, as the registrar may LIMIT, as opposed to default opt-out, where the registrar may GRANT. Here is the resulting text in -00, then -01, in 3.4.9 [4] >"Requests to retrieve information describing a registered object > MAY be granted by the registrar that currently sponsors the registered > object." Peter's suggested modification (below) allows the registry to modify the recipients of registrant-originated data, and to modify the recipients of registrar-sponsored data (usually equivalent datums). "... MAY be granted by the registrar that currently sponsors the registered object or MAY be granted by registry policy." I'd like to know what the motivating use cases are, as I don't know what Peter ment in his motivational comment by "the (registry public) retrievel". I trust we're not revisiting the whois:43 and/or whois:xx question, as that is out of scope for the R* protocol. I await further clarification. Eric