To:
Edward Lewis <Ed.Lewis@neustar.biz>
CC:
ietf-provreg@cafax.se
From:
Klaus Malorny <Klaus.Malorny@knipp.de>
Date:
Wed, 19 Oct 2005 18:08:31 +0200
In-Reply-To:
<a06200701bf7bf78d6f09@[192.168.1.101]>
Sender:
owner-ietf-provreg@cafax.se
User-Agent:
Thunderbird 1.4.1 (Windows/20051005)
Subject:
Re: [ietf-provreg] registries, XML & EPP (again)
Edward Lewis wrote: > At 18:01 +0200 10/18/05, Klaus Malorny wrote: > >> while EURid has updated their specs in a thankworthy way, this still >> has a slightly bad aftertaste, as even the namespace of the framing >> XML (i.e. the "urn:ietf:params:xml:ns:epp-1.0" namespace) was required >> to be changed due to ... > > What bothers me is that there is a second instance of a registry, in > this other case, avoiding EPP and instead building their own protocol. > I am not making a judgement call on the wisdom of prefering a home grown > effort over a standard, but rather am concerned that the standard wasn't > "good" enough for them. > > See: > http://www.ripe.net/ripe/meetings/ripe-51/presentations/pdf/ripe51-enum-epp.pdf > > > The talk is somewhat mistitled in the agenda, why DENIC didn't use EPP > for ENUM. The reason is simply that DENIC doesn't use EPP for > anything. (;)) From slide 7 onwards though is the interesting stuff - > why DENIC doesn't use EPP for anything. > > This is a cry for "use the IETF for what it is for." I would hope to > see a detailed critique outlining why EPP is failing to meet the needs > of the community come in, and not a feeling that EPP should be shunned. > The definition of EPP by the IETF is not to be taken as a royal edict. > It should be the collective wit and wisdom of us wise guys. Hi Edward, being a member of the technical advisory committee of DENIC, I strongly supported the decision of DENIC to implement its own protocol and not EPP, and I can explain why. At the time the decision was made, EPP was still in the draft state, and the underlying model of EPP was just incompatible with the DENIC model, most notably the inability to deal with a "hostless" structure. In following drafts EPP became more flexible regarding this (via the hostAttr element). However, still today there are quite a lot differences in the models, starting at additional contact fields, multiple registrant contacts per domain, different transfer mechanisms and requirements and special commands, like the transit or changeholder commands. A possible EPP solution would be either to adopt the generic EPP model or to build a large set of extensions for EPP to allow the representation of DENIC's model as EPP. The first case would not only have had a large impact to DENIC itself, but to all DENIC registrars as well. Large amounts of money would have been burnt for the developments at DENIC and the registrars, and the transition from the DENIC model to the EPP model would not have been an easy one. The other solution would have had the clear disadvantage that every single XML element added as an extension of EPP diminishes the benefit of EPP, namely to have a generic interface that can be used with a generic toolkit and a generic registrar software. Any solution in-between merely inherits the disadvantages and not the advantages of both solutions. Sorry to say that, but in the beginning of the provreg working group Scott and others of the gTLD faction were just too confined to the Verisign model and fully ignored the ccTLDs, their needs and their models. IMHO this discouraged many ccTLDs from participating. EPP is indeed "extensible", but in some aspects just not at all. While I think it is reasonable that existing registries go their own way in dealing with registry protocols for their own benefit and the benefit of their registrars and customers, I believe that new registries should adopt EPP with as few extensions as possible and with a registry model as compatible as possible, even though I do not consider myself as a big friend of EPP. Therefore I am disappointed that EURid did not manage to provide a cleaner EPP implementation within two years or so, but using the very same concepts of the .be registry. regards, Klaus ___________________________________________________________________________ | | | knipp | Knipp Medien und Kommunikation GmbH ------- Technologiepark Martin-Schmeißer-Weg 9 Dipl. Inf. Klaus Malorny 44227 Dortmund Klaus.Malorny@knipp.de Tel. +49 231 9703 0