Re: [ga] whois: issues with uniformity

[Vittorio Bertola <vb@bertola.eu.org>]

On Tue, 24 Dec 2002 13:41:41 +0100, you wrote:

>This was hotly debated in the CRISP WG (not only on the basis of
>privacy but also because of the cost it will imply for the registries,
>whose machines would have to act as benevolent servers for this
>task). The standard IETF reply is "The requirment is on the protocol,
>not on the actual registries. In other words, the service MUST be
>defined, but the registries MAY choose not to implement it."

Which may be even worse, because then you might start to rely on this
service, and then you will discover that it's broken in some TLDs that
won't support it.

>> "privileged" so that the access can be restricted to privileged
>> accounts (which, I hope, will only be released to law enforcement
>> agencies),
>
>No, priviledged will typically mean people who pay. Do not forget that
>there is an actual market for cross-registries whois searches (I know
>people who make a living from it).

And I guess this market would disappear if cross-registry searches
would become trivial.

Anyway, as an individual it really makes me feel bad that the
distinction between someone who can have access to all my personal
data and someone who can't is how much will he pay. 

Moreover, this is clearly illegal, at least in Europe. When
registering a domain, nobody ever asked my agreement in publishing my
personal data to the whole world, nor can it be reasonably said that
it is necessary for the DNS to work.

>> be written by the non-technical customer, with techies acting just as
>> support to translate the customer's needs into technically meaningful
>> language. In other words, this document (differently from the rest of
>> this WG's work) should be discussed in ICANN, not in the IETF. 
>
>I disagree here. Why would the ICANN specify the requirments for the
>future information service of .fr? The ICANN has no rights to discuss
>the whois issues in domains outside of the gTLD it manages. 

Then, why should the IETF do it? If you are talking about the actual
policy choices of any ccTLD, nobody outside the country has the right
to impose anything - at most, it may ask and give good reasons for the
request, so that people who run the ccTLD are likely to consider it
reasonable and adhere to the request.

But we are talking about requirements for a new protocol, which will
necessarily be unique all over the Internet. You can't specify a new
Internet protocol at the country level - you need to do it at the
global level. But if this new global protocol doesn't incorporate the
ability to support the policy choices you might want to make at the
country level, the ccTLDs will be prevented from adopting them,
because the protocol won't support them. (Or they will have to choose
not to support the new official Internet "WHOIS-2" protocol, which is
never a nice choice to do.)

This is why requirements for new global protocols (protocols, not
actual implementations) have to be developed at a global policy level.
If they are developed at a local policy level, they won't be global
protocols - each country will start to talk its flavour of WHOIS, and
I don't think this would be a good result. But if they are developed
at the global *technical* level, they will lack proper policy
considerations, and thus they will be unable to support all the
different policies that the countries want to implement. 

So the global policy level in this field (ICANN) should act as a
collector of all the different policies that the g- and cc-TLDs want
to implement locally, and ensure that new protocols are designed so
that all these policies are supported by the protocol. It should not
decide which types of policy are to be included in the protocol and
which ones are not - it should simply ensure that all reasonable
requests for features of the new protocol are included in its
specifications.
-- 
vb.            [Vittorio Bertola - v.bertola [a] bertola.eu.org]<---
-------------------> http://bertola.eu.org/ <-----------------------
--
This message was passed to you via the ga@dnso.org list.
Send mail to majordomo@dnso.org to unsubscribe
("unsubscribe ga" in the body of the message).
Archives at http://www.dnso.org/archives.html