Re: you have missed the point (was RE: [ga] UDRP dead?...)

[Karl Auerbach <karl@CaveBear.com>]

On Tue, 20 May 2003, Bret Fausett wrote:

> So how would you grade the wildcard issue? Is there any reason that end
> users should care? (I`m hard-pressed for a good reason to care.)

Here's how I see it:

1. If search engines or censorware engines start to accumulate a portfolio
of names that they think are registered but are really simply wildcarded
the names might be labeled (possibly permanently) as being marketing hype
(or, should a cracker take over the redirected-to site) be placed on a
censor-ware banned list.  In other words those companies that use wildcard
services are running a risk of damaging their inventory of unsold names,
rendering them either of less value or flat out unmarketable.

(I find it incredible to see an otherwise rational company risk
destruction or damage to what amounts to its entire business inventory.  
One set of bad circumstances could render NewLevel's entire DNS asset in
unsold names in .biz or .us to worthless cinders.  And for what?  A test
that could otherwise be performed in an off-line, safe sandbox?)

2. A customer who buys a name that was wild carded will discover that 
there is a period of time - at least a week, perhaps much longer, in which 
the cached material from the wildcard days lingers on.  This damages the 
front-end usability of the name, i.e. it reduces the marketability and 
market value of those unsold names because they will be encumbered with a 
history for a period of time.

3. Search engines (such as google) will become gorged on material from 
wildcarded sites - this will simply be a waste of space and a nuisance for 
as long as that material clogs the search engines' databases.

4. Those who are troubleshooting the network may find life harder because 
now virtually every name will appear at first glance to be a valid, 
registered, operational name.

5. The balance of control over who gets to handle typos will move from the 
user's end to the DNS operator.   This reduces the control the user has to 
control his/her own internet experience.  In addition, it will tend to 
clog web browser caches with even more useless garbage.

6. The cost of informing the user that he/she has mis-entered a name will
increase from today's total of two UDP packets (one UDP query packet and
one UDP response packet) to at least twice that (one packet pair to query
the TLD servers, one packet pair to query the
DOES-NOT-EXIST.default-redirect.biz. server, then at least one TCP
session, and possibly many TCP sessions, each involving at typicially 8 or
more packets, to fetch a web page that is useless to the user.  In other 
words, the traffic costs borne by users and ISPs will increase by an order 
of magnitude or more.  This doesn't include the user's wasted time while 
all of this occurs.

As I look forward to internationalized domain names I see the rate of 
typos increasing, thus multiplying the effects described above.

		--karl--







--
This message was passed to you via the ga@dnso.org list.
Send mail to majordomo@dnso.org to unsubscribe
("unsubscribe ga" in the body of the message).
Archives at http://www.dnso.org/archives.html