Shared TLD Daily Digest, Aug 01, 1996

-> .com
     by "Richard J. Sexton" 
-> Re: .com
     by Michael Dillon 
-> Re: .com
     by perry@piermont.com
-> Re: .com
     by "Richard J. Sexton" 
-> Re: .com
     by perry@piermont.com
-> Re: .com
     by perry@piermont.com
-> Re: .com
     by Michael Dillon 
-> Re: .com
     by Keith Winstein 
-> Re: .com
     by Michael Dillon 
-> Re: .com
     by Simon Higgs 
-> Re: .com
     by perry@piermont.com
-> Re: .com
     by Michael Dillon 
-> Re: .com
     by Michael Dillon 
-> Re: .com
     by perry@piermont.com
-> Re: .com
     by perry@piermont.com
-> Re: .com
     by Dan Busarow 
-> Re: .com
     by Michael Dillon 
-> Re: .com
     by Simon Higgs 


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

Date: 31 Jul 1996 12:44:49 -0700
From: "Richard J. Sexton" 
Subject: .com


Newtwork Dissolutions is going to go public.

It was bad enough they charged $50 per name to
register into the dns; this severely limits the abily
of people to make informaiton freely available
in an organized basis using the dns as a named
calulus to facilitate this.

Is there any consensus whether .com should be a shared
domain?


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

Date: 31 Jul 1996 12:58:06 -0700
From: Michael Dillon 
Subject: Re: .com

On Wed, 31 Jul 1996, Richard J. Sexton wrote:

> Newtwork Dissolutions is going to go public.

What makes you think this? There is nothing at www.netsol.com or
www.saic.com about this.

> It was bad enough they charged $50 per name to
> register into the dns; this severely limits the abily
> of people to make informaiton freely available
> in an organized basis using the dns as a named
> calulus to facilitate this.

What's a calulus? Why can't you just give away a zillion third level
domains of vrx.net?

> Is there any consensus whether .com should be a shared
> domain?

Consensus is irrelevant. The NSF and NSI own .com. Period.

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



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

Date: 31 Jul 1996 13:20:41 -0700
From: perry@piermont.com
Subject: Re: .com


"Richard J. Sexton" writes:
> Is there any consensus whether .com should be a shared
> domain?

I definitely think it can and should be.

Perry


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

Date: 31 Jul 1996 13:23:33 -0700
From: "Richard J. Sexton" 
Subject: Re: .com

> > Newtwork Dissolutions is going to go public.
>
> What makes you think this? There is nothing at www.netsol.com or
> www.saic.com about this.
>

So what. It's true. It's ben verifired.

> > It was bad enough they charged $50 per name to
> > register into the dns; this severely limits the abily
> > of people to make informaiton freely available
> > in an organized basis using the dns as a named
> > calulus to facilitate this.
>
> What's a calulus? Why can't you just give away a zillion third level
> domains of vrx.net?

There are three levels to automation. The ability to manipulate somthing
the abilty to manipulate something by numbr and the ability
to manipulate something bya named calculus - a nameing system.

There is a Chinese proverb: "The first step in organizing something
is to name it properly".

The mandate of thr WWW is "The embodiment of human information,
online". As such, thre *name* is important. It has to be easy
to rememeber and easy to type.

All information online represents somewhat of an indexing problem.
Alta Vista helps, but at somepoint, there will be so much informaiton
that it won't be *much* help.

At this point we revert to the DNS. Ideally, you should be able to
type in http://www.subject.com and get information on that subject
whether it's tractors or tomatoes. Poeple spend an inordinate
amount of time putting up unformation for free for the general
betterment of humanity - this is about more than just selling
websites, which I do by neccessity, this is about makig a usable
worls library. It does us all no good to build a free world library
and then discover somebody y owns the card catalogues.

> > Is there any consensus whether .com should be a shared
> > domain?
>
> Consensus is irrelevant. The NSF and NSI own .com. Period.

I don't believe thats true. Is this your opinion or
is this a verifiable fact ?

I was under the opinion that IANA is the guardian of .com,
and frankly I've been a bit disappointed at the way they've been
taking care of it lately.



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

Date: 31 Jul 1996 13:24:20 -0700
From: perry@piermont.com
Subject: Re: .com


Michael Dillon writes:
> > Is there any consensus whether .com should be a shared
> > domain?
>
> Consensus is irrelevant. The NSF and NSI own .com. Period.

Actually, technically this is untrue. The IANA delegated .com, but
there is no evidence that this delegation was irrevokable, or that the
NSF would not agree to making the thing shared without a fight if the
IANA had sufficient community backing for the decision.

Perry



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

Date: 31 Jul 1996 13:25:50 -0700
From: perry@piermont.com
Subject: Re: .com


Michael Dillon writes:
> > Is there any consensus whether .com should be a shared
> > domain?
>
> Consensus is irrelevant. The NSF and NSI own .com. Period.

Actually, technically this is untrue. The IANA delegated .com, but
there is no evidence that this delegation was irrevokable, or that the
NSF would not agree to making the thing shared without a fight if the
IANA had sufficient community backing for the decision.

Perry



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

Date: 31 Jul 1996 14:16:15 -0700
From: Michael Dillon 
Subject: Re: .com

On Wed, 31 Jul 1996, Richard J. Sexton wrote:

> > > in an organized basis using the dns as a named
> > > calulus to facilitate this.
> >
> > What's a calulus? Why can't you just give away a zillion third level
> > domains of vrx.net?
>
> There are three levels to automation. The ability to manipulate somthing
> the abilty to manipulate something by numbr and the ability
> to manipulate something bya named calculus - a nameing system.
>
> There is a Chinese proverb: "The first step in organizing something
> is to name it properly".
>
> The mandate of thr WWW is "The embodiment of human information,
> online". As such, thre *name* is important. It has to be easy
> to rememeber and easy to type.

Wrong. Set up virtual domains for all this info. Then set up a decent WWW
server like Apache to serve the virtual domains. Use the following
directory structure where VD stands for Virtual Domain:

   /html---
           |
           /VD1---
           |      |
           |      index.html
           |      |
           |      /subdir1---
           |      |          |
           |      |          robots.txt
           |      |          more.html
           |      |          andmore.html
           |      |          |
           /VD2---
           |      |
           |      index.html
           |      |
           |      /subdir1---
           |      |          |
           |      |          robots.txt
           |      |          more.html

Now, each virtual domain has it's own home directory. This directory
contains only the index.html document (well, maybe some other stuff like
a site map and table of contents). The bulk of the info is stored in
subdirectories which are protected by robots.txt so that web crawlers do
not index them. This focusses the webcrawlers on the index.html and
similar files. Therefore, you do a careful job of crafting that index.html
file with keywords and keyphrases to ensure that the site is "named
properly" and that it will show up on all searches that are relevant but
hopefully it will not turn up as a red herring too often because of
extraneous prose. Make sure that you use the first few words of the
index.html (which show up on the web search results) carefully to indicate
the true nature of the site's contents. Also use the META tags documented
at http://www.altavista.digital.com/cgi-bin/query?pg=ah&what=web

> All information online represents somewhat of an indexing problem.
> Alta Vista helps, but at somepoint, there will be so much informaiton
> that it won't be *much* help.

Two things will happen. One is that people will carefully craft their
sites as I described to work with the search engines. The other is that
the search engines will develop more sophisticated features especially in
their search query languages.

> At this point we revert to the DNS.

No we revert to Yahoo.

> Ideally, you should be able to
> type in http://www.subject.com and get information on that subject
> whether it's tractors or tomatoes.

You get me the capital funding and we'll set up a site that does just that
with a URL like http://www.something.com?subject
I think we can train people to do that. I think that with $2 million
up-front capital we can get this good enough to be useable and attract
advertisers.

> > Consensus is irrelevant. The NSF and NSI own .com. Period.
>
> I don't believe thats true. Is this your opinion or
> is this a verifiable fact ?

There is an RFC which talks a bit about this, but documents from NSI, NSF
and FNC all indicate that they believe this to be the case and I don't
think it will be easy to change their minds. In particular, check the FNC
www pages at www.fnc.gov where you can read the minutes of the FNCAC
meetings at which they decided to let NSI cahrge $50 fees.

> I was under the opinion that IANA is the guardian of .com,
> and frankly I've been a bit disappointed at the way they've been
> taking care of it lately.

I think IANA is disappointed at the way NSF, FNC and NSI just went ahead
and changed things without asking IANA's approval. But that's just my
personal opinion.

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



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

Date: 31 Jul 1996 14:18:57 -0700
From: Keith Winstein 
Subject: Re: .com

On Wed, 31 Jul 1996, Michael Dillon wrote:

> Consensus is irrelevant. The NSF and NSI own .com. Period.

Not to quibble, but NSF has nothing to do with com. The IANA directly
delegates com., net., and org. to the InterNIC. They do, however,
delegate edu. to the NSF, who delegates it to the InterNIC.

Keith


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

Date: 31 Jul 1996 14:23:57 -0700
From: Michael Dillon 
Subject: Re: .com

On Wed, 31 Jul 1996, Perry E. Metzger wrote:

> Michael Dillon writes:
> > > Is there any consensus whether .com should be a shared
> > > domain?
> >
> > Consensus is irrelevant. The NSF and NSI own .com. Period.
>
> Actually, technically this is untrue. The IANA delegated .com, but
> there is no evidence that this delegation was irrevokable, or that the
> NSF would not agree to making the thing shared without a fight if the
> IANA had sufficient community backing for the decision.

"Sufficient community backing" is an interesting term. First of all, if
IANA creates new iTLD's with sufficient community backing that erodes
NSI's and NSF's position because .COM is not as close to a monopoly as it
was. At this point, if we have sufficient community backing for a set of
protocols and tools which allow TLD sharing and if IANA's new iTLD process
has a built in review date regarding shared TLD's that falls just before
the expiry of NSF's cooperative agreement, then sufficient community
backing may indeed have some influence over the matter.

But right now the whole thing is a pig in a poke.

Step one: generate some movement, any movement.

Step two: steer that movement towards the goal of Shangri-La.

Step three: start repairing the potholes and washed-out bridges that we
            never noticed from way back there.

Step four: settle down in an earthly paradise because now we find that
           Shangri-La doesn't exist and besides, the earthly paradise
           is actually much nicer than we thought it would be.

Step five: imagine what the social anthropologists of the 33rd century
           will think as they explore the archives of this mailing list
           in their quest to better understand the origins of human
           society. Everybody wave to the nice anthropologists :-)

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



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

Date: 31 Jul 1996 17:40:36 -0700
From: Simon Higgs 
Subject: Re: .com

At 12:53 PM -0700 7/31/96, Michael Dillon wrote:

>> Is there any consensus whether .com should be a shared
>> domain?
>
>Consensus is irrelevant. The NSF and NSI own .com. Period.
>

That's a nice fantasy to spread to SAIC shareholders. .COM comes up for
re-delegation by IANA at the end of the current NSF contract. At that point
other registries may be delegated authority to accept registrations. That
was clearly explained by Bill Manning in the IANA conference room at the
midday meeting today.

For someone who does so much research on this subject I'm a bit dissapointed.


_____S_i_m_o_n___H_i_g_g_s______________________H_i_g_g_s___A_m_e_r_i_c_a_____
... "I'm fine - it's the others" .............. President/CEO ................
_____e-mail:  ________________  ______
....  .....




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

Date: 31 Jul 1996 17:59:39 -0700
From: perry@piermont.com
Subject: Re: .com


Simon Higgs writes:
> >Consensus is irrelevant. The NSF and NSI own .com. Period.
>
> That's a nice fantasy to spread to SAIC shareholders. .COM comes up for
> re-delegation by IANA at the end of the current NSF contract. At that point
> other registries may be delegated authority to accept registrations. That
> was clearly explained by Bill Manning in the IANA conference room at the
> midday meeting today.

This does indeed appear to be the case.

Given that, it is not irrelevant for us to discuss the question of
whether people generally feel .COM should be run as a shared TLD.

Perry


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

Date: 31 Jul 1996 18:29:38 -0700
From: Michael Dillon 
Subject: Re: .com

On Wed, 31 Jul 1996, Simon Higgs wrote:

> At 12:53 PM -0700 7/31/96, Michael Dillon wrote:
>
> >> Is there any consensus whether .com should be a shared
> >> domain?
> >
> >Consensus is irrelevant. The NSF and NSI own .com. Period.

> That's a nice fantasy to spread to SAIC shareholders. .COM comes up for
> re-delegation by IANA at the end of the current NSF contract. At that point
> other registries may be delegated authority to accept registrations. That
> was clearly explained by Bill Manning in the IANA conference room at the
> midday meeting today.
>
> For someone who does so much research on this subject I'm a bit dissapointed.

You say that Bill Manning said "may be delegated". I take that to mean
that the future is uncertain. Today NSI owns .COM. Yes, this may change
once the contract comes up for renewal in 1998, but that is the future,
not today. Right now, today, NSI owns .COM. For the near term they will
continue to own .COM. Once they cease to own .COM by virtue of the NSF
cooperative agreement, the whole question will come up of how the
authority for .COM is to be transferred because by that time I am almost
certain that there will be instances of registries having sold their
domains to other more successful registries. At that time, NSI will argue
that it should not be penalized due to grandfathering, i.e. IANA should
not just take the domain and hand it to someone else for free. When new
iTLD's are created, they are essentially handed over for free similar to
pioneer land grants. .COM is not raw unbroken prairie, it is developped
urban land and it needs to be handled with that in consideration.

I think it is wise of IANA to not try to solve the NSI and .COM problem
today but first to see if it is indeed viable to have multiple competing
iTLD registries. The problem may solve itself with a mass exodus from
.COM. It is also a different class of problem to have an NSI of two years
hence with $10 million gross annual income amongst a group of registries
with 7 others over $10 million gross annual income.

Law is a lot like network protocols. It is easy to say how things should
be just as it is easy to say that we should break up data into packets for
transmission over a wire. But the devil is in the details as you would
know if you read all the technical materials about packet-based network
protocols such as X.25, frame relay, TCP/IP and ATM. Layer upon layer upon
layer of complexity is revealed yet at any layer it is possible to
abstract the prootocls and describe them in simple terms. Same thing with
making policy or making law. Grand plans do not work. You cannot solve all
problems in one fell swoop. It takes time, study, experimentation and hard
work to develop working policies just as it takes all those same things to
create working protocols.

In any single message to this list all of us must focus on simple
abtractions that we hope accurately describe a portion of the problem in a
way that leads to better understanding and to a better design for our
policies. But it is as much a mistake to assume any one person has *THE*
solution to the entire problem as it is to assume that any one person is
totally clueless and has nothing to contribute. We are all mere mortals
here.

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



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

Date: 31 Jul 1996 18:56:38 -0700
From: Michael Dillon 
Subject: Re: .com

On Wed, 31 Jul 1996, Perry E. Metzger wrote:

> Given that, it is not irrelevant for us to discuss the question of
> whether people generally feel .COM should be run as a shared TLD.

Opinions, eh?

I think all TLD's should *EVENTUALLY* be run as shared TLD's. I'd like to
see that possibility recognized in any IANA RFC in that they extend
ownership for only a limited time, no more than two years, and that any
new registries go in with their eyes open knowing that they *MUST* share
the TLD's after their term of exclusive use. Ideally the term of exclusive
use would end the same day as the NSF/NSI coop agreement as a subtle hint.
Of course all the hoopla that happens as the registries shift to shared
TLD's will make it very hard for NSF or NSI or anyone to continue .COM as
anything but a shared TLD at that time.

And the fact of the matter is that we have NO protocols and NO running
code to handle shared registries today. So if anyone here *SERIOUSLY*
wants to see shared TLD's then the first thing to do is to write a charter
for a WG to design those protocols and get it accepted by the IETF. It
may, in fact, be able to fit into another existing WG so maybe the first
thing is to review existing WG descriptions. Note that this is not just an
international TLD thing, but that there may well be national ISO 3166
NIC's that would like to have agents dealing with all the customer
relations side of registering domains and thus they could use all the same
technology. In fact, France and the UK already are moving in this
direction to some extent so make sure to get comments on a possible WG
charter from people at RIPE and APNIC.

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



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

Date: 31 Jul 1996 19:11:12 -0700
From: perry@piermont.com
Subject: Re: .com


Kent Crispin writes:
> Personally, giving Eugene and Karl a shot at .COM is preferable
> to a hasty burst of new TLDs.

Hey, if Karl and Eugene want to run a registry under .COM and .COM is
run shared, hell, why not?

Perry


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

Date: 31 Jul 1996 19:15:48 -0700
From: perry@piermont.com
Subject: Re: .com


Michael Dillon writes:
> And the fact of the matter is that we have NO protocols and NO running
> code to handle shared registries today.

True enough, but neither is a serious challenge. Transaction systems
are bothersome to write but are very well understood.

> So if anyone here *SERIOUSLY* wants to see shared TLD's then the
> first thing to do is to write a charter for a WG to design those
> protocols and get it accepted by the IETF.

I believe that this is what we are doing here.

Perry


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

Date: 31 Jul 1996 19:55:06 -0700
From: Dan Busarow 
Subject: Re: .com

On Wed, 31 Jul 1996, Perry E. Metzger wrote:
> "Richard J. Sexton" writes:
> > Is there any consensus whether .com should be a shared
> > domain?
>
> I definitely think it can and should be.

Absolutely, as soon as the current agreement expires it should be opened
up.

Dan
- --
 Dan Busarow                                                    714 443 4172
 DPC Systems                                                  dan@dpcsys.com
 Dana Point, California      83 09 EF 59 E0 11 89 B4 8D 09 DB FD E1 DD 0C 82



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

Date: 31 Jul 1996 22:55:39 -0700
From: Michael Dillon 
Subject: Re: .com

On Wed, 31 Jul 1996, Perry E. Metzger wrote:

>
> Michael Dillon writes:
> > And the fact of the matter is that we have NO protocols and NO running
> > code to handle shared registries today.
>
> True enough, but neither is a serious challenge. Transaction systems
> are bothersome to write but are very well understood.

Agreed. It is more a matter of getting the requirements specified, and
getting the details agreed upon.

> > So if anyone here *SERIOUSLY* wants to see shared TLD's then the
> > first thing to do is to write a charter for a WG to design those
> > protocols and get it accepted by the IETF.
>
> I believe that this is what we are doing here.

So far I haven't seen a whole bunch of discussion about the text of a
charter or the timeline. You interested in doing that?

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



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

Date: 31 Jul 1996 23:54:08 -0700
From: Simon Higgs 
Subject: Re: .com

At 10:50 PM -0700 7/31/96, Michael Dillon wrote:

>So far I haven't seen a whole bunch of discussion about the text of a
>charter or the timeline. You interested in doing that?
>

I posted the start of the charter thread on Monday with "CHARTER" in the
subject line in big ol' capital letters.

Do I need to repost it?


_____S_i_m_o_n___H_i_g_g_s______________________H_i_g_g_s___A_m_e_r_i_c_a_____
... "I'm fine - it's the others" .............. President/CEO ................
_____e-mail:  ________________  ______
....  .....