Shared TLD Daily Digest, Aug 15, 1996 - Part 2

>
> Wouldnt it make sense to resolve this question first,
> or have we settled up a "some are private, some
> are shared" notion?

Whether all new TLDs are shared, or some are shared, we still to work
out the issues of how sharing would be done.  Even if there were *NO*
new TLDs, and it was decided to just turn .com into a shared registry,
we would still need to describe how to do it.  That is what I see as
the primary task of the WG.  I think it would be disingenuous and even
dishonest to say nothing at all about shared vs non-shared, but I
hope we can keep that to a discussion of the objective differences.

I don't think I have seen anyone at all say that there couldn't be
any place at all for shared registries -- the strongest statement in
that direction I have seen actually came from Jon Postel, who said he
couldn't see a business model for it.

In any case, the second output for the WG is an actual test.  We
would learn a lot from a real test.

- --
Kent Crispin				"No reason to get excited",
kent@songbird.com,kc@llnl.gov		the thief he kindly spoke...
PGP fingerprint:   B6 04 CC 30 9E DE CD FE  6A 04 90 BB 26 77 4A 5E


----------------------------------------------------------------------

Date: 14 Aug 1996 22:40:30 -0700
From: Kent Crispin 
Subject: Re: revised draft WG charter

John R Levine allegedly said:
>
> >Here is yet another revised WG charter.  The revisions were suggested
> >by Fred Baker.  Comments please.
>
> Looks pretty good, thanks for doing the work to put it together.
>
> >First, the primary underlying problem motivating change in the TLD
> >structure is resolution of trademark and possibly intellectual
> >property conflicts in domain names.
>
> Is this really true?  Whether or not it is, I'm not sure it needs to be in
> the draft.  Can you say what you're thinking about WRT shared domains?  I
> can't off hand see any difference from unshared for this particular issue.

This was added to address a specific point brought up by Fred Baker.
I may not have worded it well, but he was passing on a concern that
the creation of new TLDs has an impact on trademark issues, and it
was desired by the Area Directors that any WG in this area explicitly
address this aspect.  I actually think the issue of shared TLDs is
almost totally orthogonal to trademark issues, but I am certainly
willing to address that point in the RFC

> Also, some of us think that shared TLDs would make a lot of sense even if
> there are no new TLDs, since .COM, .ORG, and .NET would be excellent
> candidates for sharing when the Internic's current contract is up, and some
> ISO TLDs such as .US might be as well.
>
> How about adjusting the wording to allow for this possibility, or at least to
> clarify that adding new TLDs (other than the experimental .SHARED) needn't
> be a prerequisite to having shared TLDs.

I am in total agreement with you here.  I will try to add something
to this effect.  Thanks for bringing it up.

- --
Kent Crispin				"No reason to get excited",
kent@songbird.com,kc@llnl.gov		the thief he kindly spoke...
PGP fingerprint:   B6 04 CC 30 9E DE CD FE  6A 04 90 BB 26 77 4A 5E


----------------------------------------------------------------------

Date: 14 Aug 1996 23:10:23 -0700
From: Michael Dillon 
Subject: Re: revised draft WG charter

On Wed, 14 Aug 1996, Kent Crispin wrote:

> > >First, the primary underlying problem motivating change in the TLD
> > >structure is resolution of trademark and possibly intellectual
> > >property conflicts in domain names.

> address this aspect.  I actually think the issue of shared TLDs is
> almost totally orthogonal to trademark issues, but I am certainly
> willing to address that point in the RFC

Then perhaps this would be better:

     Since a primary motivation for change in the TLD
     structure is resolution of trademark and possibly
     intellectual property conflicts in domain names
     any RFC's from this WG will explicitly indicate
     what effect, if any, they might have on those concerns.

> I am in total agreement with you here.  I will try to add something
> to this effect.  Thanks for bringing it up.

Perhaps it is best to skirt the mention of .COM et al. and just point out
that existing TLD's may wish to use the sharing protocols developped by
the WG so transition issues will be adressed, i.e. transition from
one registry to multiple registries.

In larger countries such as Russia, China, Australia, Brazil, they may
even find a use for such protocols with their national domains. For
instance, if domain name registration becomes a government service why
couldn't I go to my local government agent's office and apply for a
.BC.CA domain just like I now submit a name search for a new corporation?
This is an example of a scaling issue that may not make sense in the small
global Internet of today but will be quite relevant ten years from now
when there are 50 million domains or so.

Michael Dillon                   -               ISP & Internet Consulting
Memra Software Inc.              -                  Fax: +1-604-546-3049
http://www.memra.com             -               E-mail: michael@memra.com



----------------------------------------------------------------------

Date: 14 Aug 1996 23:10:36 -0700
From: Kent Crispin 
Subject: Re: new draft WG charter

Michael Dillon allegedly said:
>
> On Tue, 13 Aug 1996, Kent Crispin wrote:
>
> > 	o TBD
> > 	  (Someone from IANA?)
>
> Before you volunteer one of the four people who work part time on the IANA
> activities, you really should ask them. I don't see any need for IANA
> people to chair a shared TLD protocol WG and I think they will have more
> important things to deal with in the next little while as new TLD's are
> deployed. It wouldn't hurt to ask someone like Perry Metzger or Paul Vixie
> if they would be interested or if they could suggest people for the
> position.

These were the very next two people on my list, in fact.

> > Internet Area Director(s)
> >
> > 	o Frank Kastenholz 
> > 	o Jeffrey Burgan 
>
> Have you asked either of these guys? The shared TLD stuff sure seems like
> OPS area stuff to me.

As Fred Baker suggested -- that's been changed in the revision I sent
out shortly after this one.

> > Description of Working Group
> >
> > The Shared Top Level Domains Working Group is concerned with the
> > technical and logistic requirements of creating shared domain name
>
> shared access to domain name ...

OK

> > registration databases, and the administration of delegated top level 
> > domains by multiple domain name registries.  The motivation for this
>
> multiple registration agents.

OK

>
> > concern is to minimize centralized management of all components of
> > the name space.
>
> > Since the proposals of this WG would directly impact the workings of
> > the IANA, it is recognized that the IANA must actively participate.
>
> Drop this bit. Either they will or they won't. In fact, I disagree that
> this WG would *directly* impact the workings of IANA. Perhaps IANA will
> choose to do things differently in the future due to the deployment of
> this WG's work, but IMHO that is rather a indirect impact and could be
> said of many WG's.

I have given a fair amount of thought to this, and I don't think the
involvement with IANA can be ignored.  Part of the proposal will be to
propose policies as to how IANA will delegate control to TLD
registries.  And one very real possiblity is that the IANA would be the
entity that actually runs the coordination database.  So I don't believe the
issue of IANA involvement can be responsibly avoided.  That's why, of
course,  I also think an IANA chairperson would be very valuable.

>
> > The primary products of this WG are three:  First the STLD Draft, and 
> > second,
>
> define this draft more clearly. IMHO something like this:
>
>      First, the definition of a protocol for shared administrative
>      access to domain name registry databases capable of reserving
>      names, registering names and modifying name records in a
>      secure authenticated manner.
>
> > a test implemtation of the ideas using the .SHARED TLD as a
> > test,
>
> Just say that it is a standards track protocol which implies two
> implementations.
>
> > and three, a revised STLD draft incorporating the lessons from
> > the experiment, which should become an RFC
>
> This is also implied by "standards track". Rather than stating the obvious
> here, try to be more clear about why the protocol is needed, what the
> requirements are, etc.

Perhaps.  I'll have to think about this one.  One model of how this
done (the "chaotic" model) doesn't fit this mold very well.

> > 	o fostering an appropriate blend of competition and
> > 	  cooperation between cohort registries
>
> good.
>
> > 	o the relationship between STLD registries and the DNS
>
> vague.
>
> > 	o the relationship between STLD registries and the IANA,
> > 	  including suggested procedures for licensing a registry,
> >           and the associated policies
>
> relationship is a bad word you know.

As Bill Murray said in Ghostbusters, "I'm fuzzy on this good/bad
thing." :-)

> And IANA has nothing to do with this
> really. They will create TLD's but unless they mandate some sharing
> timeframe this whole thing will be voluntarily done by some registries.
> Also, you are overlapping the Postel draft in this clause.

I am quite convinced that there will be potential significant impact
on the IANA.  We can debate that issue on the list.  If it turns out
the impact is minimal we can say so.

> > 	o technical issues regarding management of the various
> > 	  databases involved in running and coordinating registries.
> > 	  This includes issues like distribution of updates, locking
> > 	  of a shared coordination database, and so on
>
> Rather than a partial list, try to choose words which describe the
> complete scope of the problem. It is better to define boundaries and only
> be specific with things that are mandatory result items. Within the
> boundaries defined by the charter it is quite possible that the WG will 
> come up with new issues and/or decide that some issues currently mentioned
> are actually not relevant.

You are right.  I will just delete the examples.

> > 	o the adminstrative procedures involved in running
> > 	  a registry serving a Shared TLD
>
> This focuses too much on internal registry procedures rather than the
> administrative protocols between registries which are where it should
> focus, IMHO.

This is a wording problem.  The adminstrative procedures I am getting
at are exactly the ones necessary to coordinate with cohort registries.


> > 	o authentication and authorization issues
> >
> > 	o minimizing possible legal complexities
> >
> > 	o as much as possible, making the relationships between
> > 	  cohort registries self-regulating -- that is, minimizing
> > 	  IANA's role in regulation or dispute resolution.
>
> Again, leave out IANA. It is not a court. If IANA publishes and RFC that
> gives it arbitration rights, then so be it. If not, then leave them alone.

Once again, I disagree.  I do have mail in to IANA folks, and I will
discuss it with them.

> > Jan 1997
> > 	Submit a Shared Top Level Domain IETF Draft that outlines
> > 	one or more technical and policy solutions, along with
> > 	a fairly detailed discussion of the tradeoffs that led to
> > 	if (them).
>
> Tradeoffs is good. It helps future designers understand the genesis of the
> protocol. However I don't like "one or more". A single draft should
> promote a single solution.

There have been three overall models discussed so far, and I would
like to remain open to the possibility of supporting a minority camp,
if necessary.  I expect the review will indeed whittle things down to
one, but I would rather keep a bit friendlier environment while we
discuss things.

- --
Kent Crispin				"No reason to get excited",
kent@songbird.com,kc@llnl.gov		the thief he kindly spoke...
PGP fingerprint:   B6 04 CC 30 9E DE CD FE  6A 04 90 BB 26 77 4A 5E


----------------------------------------------------------------------

Date: 14 Aug 1996 23:33:03 -0700
From: Michael Dillon 
Subject: Re: new draft WG charter

On Wed, 14 Aug 1996, Kent Crispin wrote:

> > > Since the proposals of this WG would directly impact the workings of
> > > the IANA, it is recognized that the IANA must actively participate.
> >
> > Drop this bit. Either they will or they won't. In fact, I disagree that
> > this WG would *directly* impact the workings of IANA. Perhaps IANA will
> > choose to do things differently in the future due to the deployment of
> > this WG's work, but IMHO that is rather a indirect impact and could be
> > said of many WG's.
>
> I have given a fair amount of thought to this, and I don't think the
> involvement with IANA can be ignored.  Part of the proposal will be to
> propose policies as to how IANA will delegate control to TLD
> registries.  And one very real possiblity is that the IANA would be the 
> entity that actually runs the coordination database.  So I don't believe the
> issue of IANA involvement can be responsibly avoided.  That's why, of
> course,  I also think an IANA chairperson would be very valuable.

> > And IANA has nothing to do with this
> > really. They will create TLD's but unless they mandate some sharing
> > timeframe this whole thing will be voluntarily done by some registries.
> > Also, you are overlapping the Postel draft in this clause.
>
> I am quite convinced that there will be potential significant impact
> on the IANA.  We can debate that issue on the list.  If it turns out
> the impact is minimal we can say so.

If IANA really feels that this is tightly knitted into what they are
doing, then perhaps it would be better to get someone from there to
develop the wording re IANA. Maybe the relationship with IANA, policies
and the like should be a separate RFC from the technical design issues
regarding sharing. Or maybe not. Depends on how big an impact this has on
IANA activities which are already moving into the international venue.

> This is a wording problem.  The adminstrative procedures I am getting
> at are exactly the ones necessary to coordinate with cohort registries.

        o administrative procedures required to coordinate
          registries which share the same database

> > > Jan 1997
> > > 	Submit a Shared Top Level Domain IETF Draft that outlines
> > > 	one or more technical and policy solutions, along with
> > > 	a fairly detailed discussion of the tradeoffs that led to
> > > 	if (them).
> >
> > Tradeoffs is good. It helps future designers understand the genesis of the
> > protocol. However I don't like "one or more". A single draft should
> > promote a single solution.
>
> There have been three overall models discussed so far, and I would
> like to remain open to the possibility of supporting a minority camp,
> if necessary.  I expect the review will indeed whittle things down to
> one, but I would rather keep a bit friendlier environment while we
> discuss things.

I guess I interpret the word "draft" as meaning a document that will
become an RFC. If you are referring to a draft document that is a review
of current work then you are right because I was referring to standards
track protocol stuff. Maybe this could be made clear. In particular, do
these reviews of current work ever become RFC's? We should be a bit
careful here with names of documents and I think it helps if we explicitly
say that a milestone is to work on an informational RFC or a BCP or an FYI
or a standards track document just so people can clearly distinguish
between the multiple documents that can come out of a WG.

Michael Dillon                   -               ISP & Internet Consulting
Memra Software Inc.              -                  Fax: +1-604-546-3049
http://www.memra.com             -               E-mail: michael@memra.com