To:
ietf-provreg@cafax.se
From:
Marcel Schneider <schneider@switch.ch>
Date:
Fri, 05 Jan 2001 11:38:25 +0100
Content-ID:
<29107.978691105.1@smtp.switch.ch>
In-reply-to:
Message from "Peter Mott" <peter@2day.com> of "Fri, 05 Jan 2001 23:15:10 +1300." <NDBBLCLIJMHJGOKHMOEBKEODFHAA.peter@2day.com>
Sender:
owner-ietf-provreg@cafax.se
Subject:
Re: Definition of Registry
On Friday, 5 Jan 2001, "Peter Mott" writes: I completely agree with Peter's analysis blow. But the important fact for this group is: there are two models for registries/registrars/agents. The 'lightweight' registry model (just DB and connectivity) is mostly used in gTLD's and will continue to be (one of the reasons is that it has not the registry as a bottelneck). The 'policy-setting' registry is more common in ccTLD's and will continue to be. The RRP needs to support both models. We should discuss the reasons why it couldn't. Marcel >> it's the practice for the .fr registry >> update of the database are requested by the registrant, through >> the registrar. > We have different views of the world. Through my eyes I see domain name > registration being the execution of a legal contract between the registrar > and the registrant. The public record of that contract is the insertion of > the name into the registry database. > To determine the registrar, I simply ask the question "with whom does the > registrant have a contract for the domain name registration?" Whoever that > entity is - is always the registrar. > Not sure how .fr works, but I suspect the contract is in fact between the > registrant and your organisation. The party you refer to as registrar is > probably acting as an agent for the registrant. Your organisation is in > fact the registrar. > This may all seem to be splitting hairs, but its important stuff in my view. > Understanding relationships and roles is required before developing system > to support it, and protocol to enable system >> registrar is a provider of added value (services) for the user >> connectivity. > My definition of registrar is simply an entity accredited to enter into > domain name registration contracts with registrants. This entity may have > domain name registration as a core or supporting business activity. It is > not relevant. >> In France, registry is a monopoly model (public service, not >> registrar itself), >> registrars are in the competitive sector. > I think you will see from above that I dont agree with you. What you likely > have is a monopoly registrar (your organisation) with agents all offering > the same registration contract (your TLD policy combined with business > rules). > Such a model does not support competition in the business of domain name > registration. Although I concede that agents can wrap your service in > different ways to differentiate themselves. I suspect however the major > differentiation occurs between core service elements for the agent rather > than the service of name registration. > Please dont think I am a supporter of any particular model or have fixed > views. I have been studying this stuff in depth since 1996, own an > accredited ICANN registrar (although not operational as yet) and admin for > .tk .aq as well as registry operator for .pn > What I have been trying to highlight is that it is the identification of the > players and their relationships that is important, not so much where the > data is. If this wg proceeds with the relationship model in the draft, > thats ok with me. I just wanted to make sure people had the opportunity of > considering alternatives. > Regards > Peter Mott > Chief Enthusiast > 2day.com > -/-