Internet Info 1997 December/Internet_Info_CD-ROM_Walnut_Creek_December_1997.iso/ietf/urn/urn-archives/urn-ietf.archive.9704/000021_owner-urn-ietf _Mon Apr 21 12:35:53 1997.msg

Internet Message Format | 1997-04-30 | 8KB

Received: (from daemon@localhost)
    by services.bunyip.com (8.8.5/8.8.5) id MAA28025
    for urn-ietf-out; Mon, 21 Apr 1997 12:35:53 -0400 (EDT)
Received: from mocha.bunyip.com (mocha.Bunyip.Com [192.197.208.1])
    by services.bunyip.com (8.8.5/8.8.5) with SMTP id MAA28016
    for <[e-mail]>; Mon, 21 Apr 1997 12:35:41 -0400 (EDT)
Received: from acl.lanl.gov by mocha.bunyip.com with SMTP (5.65a/IDA-1.4.2b/CC-Guru-2b)
        id AA27761  (mail destined for [e-mail]); Mon, 21 Apr 97 12:35:39 -0400
Received: from montana (montana.acl.lanl.gov [128.165.147.143]) by acl.lanl.gov (8.7.3/8.8.5) with SMTP id KAA05633; Mon, 21 Apr 1997 10:34:00 -0600 (MDT)
Message-Id: <[e-mail]>
X-Sender: [e-mail]
X-Mailer: Windows Eudora Pro Version 3.0 (32)
Date: Mon, 21 Apr 1997 10:32:51 -0600
To: "Dirk.vanGulik" <[e-mail]>
From: "Ron Daniel, Jr." <[e-mail]>
Subject: Re: [URN] NID rambling
Cc: [e-mail]
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Sender: [e-mail]
Precedence: bulk
Reply-To: "Ron Daniel, Jr." <[e-mail]>
Errors-To: [e-mail]

At 11:00 AM 4/14/97 +0200, Dirk.vanGulik wrote:
>Just to spark off discussion...

Sorry not to have responded earlier. Now for the sparks.  :-)

>1.    The registry resides in a country with
>    a sane legal system.

Care to define "sane"?

Dirk and I have had enjoyable chats about the legal systems of the US
and the EC. However, this requirement is too hard to measure, and would not
preclude US courts getting involved. Also, since IANA is the body that has
served as the registrar of all assigned IDs for the Internet so far, I think
they need to keep the job. While the people at IANA might rather
enjoy moving to Monacco, we should probably assume they will remain in
California.

>2.    IANA _assigns_ NID's upon request; rather
>    than allow the requester to choose. IANA
>    'should' use a simple counter, a random
>    string, or simply time time in seconds
>    or something that the request arrives. IANA
>    'can' assign a name like ISO or ISBN if it
>    wants to.

What I think we need to establish are the rules whereby IANA can say
"you asked for 'foo' and you get it" or "you asked for 'foo' and you don't
get it".


Here are my suggested rules:

  0) Select the desired NID. "foo" will be used as an example.

  1) Is "foo" a registered URL scheme? If so, request denied.
     (We need to keep the URN NIDs and URL schemes disjoint to preclude
      future ambiguity).

  2) Is "foo" an acronym for a naming system standardized by ISO?
     (This rule requires that there be a search mechanism over existing ISO
      standards and current ISO committees to look for names).
     If not, go to question 3.
     If it is, does the requested use follow that standard?
     If not, request denied. (Deliberate conflict with ISO is stupid,
        since they are the law of the land in most countries).
     If it does follow ISO use, the name may be registered. (Adding
       the NID to any particular resolution system may require additional
       work, such as a demonstration that all vendors have an equitable
       chance at participating in the resolution system).
     If ISO has multiple naming systems with the same name, let them figure
       it out.

  3) Is "foo" the product of a international industry standards body such
     as the IEEE, SMPTE, etc.?
     (Augment "foo" with the organization name, such as urn:foo.ieee:1234)

  4) Is "foo" the product of a national industry standards body such as
     the NAM, NAB, EIA, etc., and is the use of names going to follow that
     standard?
     If so, augment the name with the name and locality of the standards
     body. (There are issues here of non-uniqueness of standards group IDs
     within a jurisdiction. NAB might be Natl. Assoc. of Broadcasters, or the
     Natl. Assoc. of Ballerinas. Not sure what to do there).

  5) Is it a namespace established by a
     commercial organization for the use of any organization following
     normal business practices?  (This is to allow namespaces such as
     DUNS (30 million identified organizations), some of the Barcode
     registries (each can accomodate a few 10s of thousands of companies),
     etc. It is intended to disallow vanity namespaces. This may be
     impossible.)
     Is there a way for organizations to assign names in the system?
     (For the case of DUNS, Dun & Bradstreet would like to identify
     certain reports they sell. That's fine, but I see no need for the IETF
     to standardize it. If however, companies can use their DUNS number
     in conjunction with some information they specify themselves, we can
     consider that.

>Soap:    I do insist that globally accepted names do
>    NOT exist; no matter what companies claim. Check
>    out the former East Germany, or NATO/OTAN. ISO/OSI
>    etc, for examples.

Sorry, Dirk. You are trying to prove a negative. Can't be done.

>3.    sub 2. A company can reject a name assigned by
>    IANA, and resubmit. There is a small fee to be
>    paid if you reject.

Being able to say "no, I am not going to use the NID you have sent me" is
useful functionality. I think we leave questions of procing policy to IANA.

>4.    Submiters pick their own name, and provide a
>    legal letter saying that that name is theirs in
>    a certain juristdiction. The NID is postfixed
>    by '/'.jurisdiction; i.e. ISO/CH or SHELL/NL.
>
>    This implies that any claims can be settled by
>    civil court between claimants within that
>    jurisdiction.

And what happens to jurisdictions when a country has a civil war and all
sides claim to be the "true" owner of the ISO 2-letter county code? More
to the point, trademarks are not unique even within one jurisdiction. A
brick company, an electrical company, and a bakerey could all trademark
the term "Acme" giving them rights to it for their product family. So,
even a letter claiming to hold a trademark is not a guarantee of uniqueness.

I think this is actually a pretty good suggestion for how to deal with
vanity name requests, although it will need some more detail. I also
think that there are some names, such as ISBN, that are not vanity names
but are indicative of existing international standards. I would *really*
like to make sure that the rules for operating the registry will allow
those to be treated as simply as possible. I don't really want to see
urn:isbn/iso/ch:1-234-5678-9, for several reasons.

>    If a company claims to have a name in more than
>    one jurisdiction it MUST supply the credentials
>    for _EACH_ jurisdiction; but will only get the
>    name in ONE. However in the other jurisdictions
>    the claim prevails.

Right. I'd augment this with a procedure where IANA notifies the holder
of a name in one jurisdiction when someone tries to register it in
another jurisdiction.

>5.    Use OIDs

I would like to see OIDs as a namespace, such as
   urn:oid:1.2.3.4.5.6.29.2.1002
Using OIDs as the NID essentially restricts URNs to being isomorphic to the
OID namespace, which seems too restrictive to me.

Later,

Ron Daniel Jr.              voice:[phone]
Advanced Computing Lab        fax:[phone]
MS B287                     email:[e-mail]
Los Alamos National Lab      http://www.acl.lanl.gov/~rdaniel
Los Alamos, NM, USA, 87545