Addressing the name speculation problem

I think there are probably several possible solutions to the land-grab
problem, and I'll suggest one or two in this note. Before I do that, I
want to reinforce the point that solving a potential problem by
crippling one of the essential technical requirements we want in the URI
namespace is almost certainly a mistake, and we should avoid it if at
all possible, and I think it IS possible to avoid it.

The Latency Approach

Larry, in a private message to me, has hinted at one way.  I'm not sure
if it is a serious proposal at this time, or more in the nature of
thinking out loud, so I'll leave explication of his ideas to him.  Lets
call it the Latency Approach.

The Toll Gate Approach

A way I've thought about but have hesitated to suggest until now is put
up a toll gate that will inhibit speculation.  We currently charge
people nominal fees for domain registrations.  Why? Because there is
infrastructure that must be supported.  The amount is so low as not to
be an impediment to name squatting, but no one seems to have run out of
clever URL phrases in any case.  URI schemes are different... We don't
expect thousands, and one might argue that short names are much more
important than with URLs, so we think we need to protect the juicy ones.
Let me note that I think it is unlikely that we'll run out of juicy ones
anytime soon, either, but I don't have a strong conviction about how
serious the problem will be, and I agree it is prudent to anticipate a
problem.

What sort of toll will retard the speculative introduction of vanity URI
schemes while not unduly penalizing well-thought-out innovation?  Some
amount of earnest work and some amount of money are both considerations.
The work part might be an internet draft or similar documentation
requirement, and that is already part of the scene.  The money part
might be a registration fee (that could in turn be used to promote more
rapid evaluation of proposals.  What would a sensible amount be?  $1,000
USD? $2,000? $5,000? It is hard to tell if this would be a significant
impediment to squatting but I think it might be.  

The Endorsement Approach

A possible policy safeguard might be to require multi-organizational
endorsement.  A required section of an ID might be an
endorsement-of-utility assertion: signed assertions by individuals from
three independent organizations asserting the recognition by the
organization that the proposed scheme meets a set of requirements and is
important to a community.  The assertions would be made by individuals
representing the organizations, assuring both individual accountability
and organizational accountability.  Endorsements of frivolous schemes
would be like pyublishing bad data in a paper.  

Please note that none of these are proposals... They are simply
responses to Dan's question...  brainstorming *policy* solutions to a
*policy* problem.  What we are facing at the moment is a *technical*
solution to a *policy* problem, and I think that this is generally a
dangerous way to enact policy.  The side effects will be unpredictable
and pernicious, and a cure, if it should be needed, is likely to be
harder to implement.

stu


-----Original Message-----
From: Dan Connolly [mailto:connolly@w3.org] 
Sent: Monday, February 14, 2005 1:54 PM
To: Weibel,Stu
Cc: uri@w3.org
Subject: RE: Duplication of provisional URI namespace tokens in
2717/8-bis

On Mon, 2005-02-14 at 13:47 -0500, Weibel,Stu wrote:
> They clearly are related, but they are different requirements and 
> distinct mechanisms for coping with them can be elaborated.

Maybe I'm just not very creative, but I don't see how.

An example of how to address the land-grab problem without allowing
duplicates would make your point much easier for me to see.

> As currently written, 2717/18 needlessly sacrifices one requirement to

> guard against danger to the other.
> 
> stu
> 
> -----Original Message-----
> From: Dan Connolly [mailto:connolly@w3.org]
> Sent: Monday, February 14, 2005 1:33 PM
> To: Weibel,Stu
> Cc: uri@w3.org
> Subject: RE: Duplication of provisional URI namespace tokens in 
> 2717/8-bis
> 
> On Mon, 2005-02-14 at 13:04 -0500, Weibel,Stu wrote:
> [...]
> > Larry Masinter raises a legitimate concern about land-grab 
> > speculation
> 
> > of URI scheme names.  This concern deserves attention, but must be 
> > divorced from the functional requirement of unique scheme tokens in 
> > the URI space.
> 
> ???
> 
> They're intimately connected. I don't think it's possible nor 
> advisable, let alone mandatory, to separate them.
> 
> The land-grab phenomenon results from scarcity.
> Uniqueness creates the scarcity.

--
Dan Connolly, W3C http://www.w3.org/People/Connolly/
D3C2 887B 0F92 6005 C541  0875 0F91 96DE 6E52 C29E

Received on Tuesday, 15 February 2005 14:07:15 UTC