[Ietf-not43] From answer to domain

[Chris Ambler]

A small bit of history - Rick Wesson published a draft a couple of years ago
with a simple XML format for WHOIS exchange. Essentially, an XML
encapsulation of all of the stuff you'd want to see in WHOIS output. I took
that draft and changed it to support what we need today. Rick took a look at
it and said, "Hey, dreg has all of this stuff - why not just use that rather
than make a whole new format?" That made sense to me, as the dreg schema
does, indeed, have everything we need.

The solution is a constrained one. We want to just emit domain information,
and we want to do it in a single transmission - no extra trips to the server
to resolve IDs or handles.

As such, the solution seems to be to emit a resultSet with an answer
containing a single domain, and an additional element with the contacts and
nameservers and such in the same resultSet.

The goal here is to create a WHOIS interchange format. Period. At Rick's
urging, we're using dreg instead of something new, simply because they'd be,
essentially, the same thing. The only difficulty here is in getting the data
into a single transmission, which is made slightly challenging because of
the normalization issue. But including the elements in the additional
element should solve this.

So the fact that we don't re-invent the wheel should mean that this makes
sense then and there. As for the future, it allows us to be open to more of
the features for this and future projects, and to more easily integrate our
WHOIS output with other, larger IRIS-based projects. No?

Christopher

-----Original Message-----
From: David Blacka [mailto:davidb at verisignlabs.com] 
Sent: Wednesday, October 15, 2003 11:05 AM
To: Chris Ambler
Cc: ietf-not43 at lists.verisignlabs.com
Subject: Re: [Ietf-not43] From answer to domain


On Wednesday, October 15, 2003, at 01:08  PM, Chris Ambler wrote:

> Heretical question: if the XSD is just a suggestion, as you intimate, 
> what
> prevents me from breaking normalization and adding the allowance that, 
> for
> example, the technicalContacts element of the domain element may 
> contain, in
> addition to an iris:entity element, a contact element?
>
> Okay, let's say I don't do that (because your objection will be, "yes, 
> that
> breaks normalization, and we'd prefer that not happen), but put the 
> actual
> element in the additional element... is there a procedure whereby I 
> can push
> my XSD back out as the "official XSD for using the dreg:domain inside 
> of an
> answer element for the purpose of registrar WHOIS transmission?" or 
> some
> such?

I might be misunderstanding this whole conversation, but it seems to me 
that you are trying to use the IRIS/DREG1 schema for some 
non-IRIS/DREG1 purpose, and that you are having some difficulties 
because of this.

Is the goal to create a WHOIS interchange format that will interoperate 
with CRISP, or is this an independent effort?  (By interoperate, I am 
thinking that there may be a desire for a registrar to use one of these 
files as the data source for an IRIS/DREG1 server).

If this is an independent effort, then you should just be using the 
DREG1 schema as a starting point.  The DREG1 schema was designed in the 
context of the IRIS architecture.  If that architecture isn't there, 
then the schema isn't going to be ideal.  In that case, you may as well 
as create your own (albeit similar) schema.

If it isn't an independent effort, then we need to make sure we 
understand how your efforts and IRIS need to interoperate.  The schemas 
are not set in stone, but we need to understand the problems before we 
can judge what changes (if any) need to be made to the schemas.

--
David Blacka    <davidb at verisignlabs.com>
Sr. Engineer    Verisign Applied Research