Re: [wg-c] Eureka?

[Kent Crispin <kent@songbird.com>]

On Fri, Aug 06, 1999 at 05:39:52PM -0400, Milton Mueller wrote:
> Kent Crispin wrote:
> 
> > there is only one name
> > space.  Either it is planned, or it is not.  It is simply incoherent
> > to say that it is planned, but that we allow free choice of names to
> > some set of participants -- that free choice undermines any
> > possibility of planning.
> 
> This is an extreme, ideological claim. All forms of public planning actually
> combine free choice with constraints and standards, otherwise they are just 
bad

Note that the above line had precisely 79 characters, and when the 
standard quoting convention of prepending a "> " is exercised, the
format of your text ends up like this:

> xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
xx
> xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
xxx
> xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
xxxx

After a couple of layers of quoting it becomes incomprehensible.

This makes it really tedious to wade through the stuff you write, 
and motivates me, at least, to ignore it.  The RFCs recommend that 
you keep your lines comfortably short of 80 characters, because, 
ancient though it is, many environments still use 80 columns by 
default.  In stead of going through my usual tedious reformatting, I 
will return the quoted text so you can see how it looks to me, along 
with a reminder of how big it is.

> plans. People are going to make choices, and if your plan doesn't take that 
into

80 characters...

> consideration it will fail.

Of course my plan will take that into consideration.  Notice what a 
pain it is to have distracting structural elements interfering with 
the flow of what you are trying to read?

[...]

> You have yet to explain what practical benefit to real Internet users accrue
s from 

83 characters...

> your so-called "planned" name space. Indeed, where is your plan?

It is quite simple, and I have mentioned it several times already:
All new TLDs must be approved through a DNSO WG chartered for that
purpose.  For example, CORE could ask the NC to approve a WG to
consider .rec.  The WG would provide the forum for the planning
aspects to be considered.  You could think of the DNSO as a planning
commission; the WGs as fact finding committees in the planning commission.

There is absolutely no doubt that this would be quite tedious at 
first, but the first few TLDs would set precedents and expose issues 
that rules could be built from, and the process would become much 
more efficient.  This process has the natural characteristic of 
picking up speed with time.

> Absent a plan, I am quite sure that the practical difference between your
> so-called "planning" and the plans and choices made by registries who are ab
le and

83 characters...

> willing to operate TLDs is not very significant.

Good.  Then you shouldn't mind my planning efforts. :-)

> You think there should be 6 gTLDs. It is no secret that your favored 6 corre
sponds

83 characters...

> to the IAHC list, minus (maybe) dot Web.

That is my favored six, 'tis true.  However, they should stand or 
fall on their merits as we go through open public debate.  I think 
there are good arguments for them, but that is a different debate.

> Now, in my model, Core offers itself as a shared registry to ICANN and says,
 "we

81 characters...

> want to operate .nom, .shop, .info." IOD and perhaps Name.space offer to ope
rate

82 characters...

> .web. another company would probably claim .arts, and so on. There is room f
or new

83 characters...

> ideas, however.
> 
> In your model, this committee says "we are authorizing .nom, .shop, .info., 
.web,

82 characters...

> .arts and .rec." Your "plan" just means: accept the plan that the gTLD-MoU p
ut

79 characters...

> together two years ago.

No committee is involved, only democratic, open, public processes.  I
would think you would be wildly in favor of that.  

My plan, as I expressed it in an early message to this list is to
authorize 6 TLDs, to be run by at least 3 distinct shared 
registries.  The registries will be non-profit, cost-recovery based, as 
far as transactions with registrars are concerned; but they can be run 
under bid by a for-profit company.

> There are indeed important differences between what you and I have in mind w
hen it

83 characters...

> comes to business models, exclusivity, and the degree of consumer and suppli
er

79 characters...

> choice. But in terms of the structure of the name space, you are not proposi
ng

79 characters...

> anything significantly different.

Glad to hear it.


> Perhaps you do have some master plan for a perfectly structured name space t
ucked

82 characters...

> away in your pocket. Perhaps it is the epitome of Cartesian rationality--it
> anticipates anything anyone could possibly want from the name space, now and
 in

80 characters...

> the foreseeable future. Guess what? By the time any "open, public process" g
ets

80 characters...

> finished with that plan, it won't look anything like what you have in mind, 
and it

83 characters...

> won't be any more rational than the supplier-driven additions that I am prop
osing.

83 characters...

Yes, it will, because the model you propose will suffer from what is
a variant of the "tragedy of the commons", just as .com/.net/.org
have.  The procedure I describe *will* allow coherent chunks of the 
name space to be planned.

> Indeed, from a market point of view it will be considerably less rational.
> 
> End users have consistently rejected planned, organized domain name spaces i
n
> favor of open, flexible ones. Most suppliers have shown little interest in t
hem.

81 characters...

> Do you not care what the people who actually pay for, use, and supply domain
 names

83 characters...

> want?

> > You have a
> > tract of land, and you say part of it will be planned development,
> > and part of it will be given to a developer to develop, free of any
> > plan.  The planned part is to be for residential units.  The
> > developer decides to build a toxic waste dump.  The two uses don't
> > work together.  In fact, the use of the land as a toxic waste dump
> > prempts the use of any of the nearby plots for most purposes.

Notice how my doubly-quoted text, above, maintains its neatness and 
visual integrity, whereas your text ends up as an almost unreadable hash?

> FYI, these are called "negative externalities." Perhaps you can explain what
> negative externalities would be created by the co-existence of shared and
> exclusive gTLDs, or the co-existence of "planned" TLDs and unplanned naming
> choices. It is hard to understand how the presence of unplanned name choices
 would

83 characters...

> generate some Internet equivalent of "toxic waste"-- something directly harm
ful to

83 characters...

> users or suppliers. Again, try to be specific and concrete. We need to progr
ess.

81 characters...

> 
> What I think you really mean--and pardon me if this is a bit cynical--

I have come to expect nothing but.

> is that you
> fear that if name choices and entrepreneurial activity is allowed, users wil
l

79 characters...

> choose the "free" names over your "planned" names. And therefore your plan w
on't

81 characters...

> work.

No, I don't fear that at all.  What I fear is that, in the domain name 
squatting tradition, we will see "entrepreneurs" grab up the good 
names, the ones that would be most useful in a structured context, 
and hold them hostage.  The model that you and Weinberg and others 
propose of essentially unlimited access to the root zone will 
inevitably result in thousands of gTLDs and private TLDs, because, 
as has been pointed out, there is almost no incremental cost to 
adding more TLDs to a registry.  Unless you adopt the stance from 
the beginning that the root zone will be planned and controlled, 
there will be no stopping endless expansion.  Please remember 
Amblers threats to sue his way into the root.  One has to be 
extremely careful about the precedents that are set, so that you 
don't give legal grounds for such actions.

-- 
Kent Crispin                               "Do good, and you'll be
kent@songbird.com                           lonesome." -- Mark Twain