IP: more on ICANN To Evaluate Internet Keywords

[David Farber <dave () farber net>]

> Date: Thu, 24 Jan 2002 11:51:16 -0800 (PST)
> Reply-To: Karl Auerbach <karl () cavebear com>
> To: farber () cis upenn edu
> From: "Karl Auerbach" <karl+dated+1012333876.154957 () cavebear com>
>
>
>
> ICANN is not "evaluating" keywords.  And keywords are not part of ICANN's
> role.
>
> Naming systems of the Internet are evolving; new directory services are
> being created.  The sensible ones are being built using DNS, if they use
> DNS at all, merely as an underlying layer that isolates the dynamics of IP
> address changes and uses DNS merely as a relatively unchanging namespace
> and one in which the interpretations that humans may chose to apply to the
> character sequences used are not of any particular importance or concern.
>
> How DNS is used by users or applications is not part of ICANN's job.
> ICANN's role is merely to ensure that DNS runs smoothly and reliably.
>
> It would be unwise for ICANN to expand its role and to try to regulate
> beyond DNS itself.  (But that horse may have been out of the barn for a
> long time - ICANN has already gone far beyond mere "technical"
> coordination of DNS - one has to be naive indeed to believe that the UDRP,
> ICANN's domain name dispute policy, has any tie to "technical
> coordination.")
>
> I, personally, believe that the edge of DNS is the point in the software
> at which a given computer makes its final decision to emit a RFC compliant
> packet aimed at port 53 on a DNS server.  Anything that occurs before the
> software's decision to use DNS is, to my way of thinking, simply not part
> of DNS itself and is not part of ICANN's job.
>
> One particularly interesting instance of this is the path from a web
> browser's "address bar" to whatever pops up inside the display window.  It
> is a long and winding path - and it is becoming a target of control (and
> right now on Windows platforms, Microsoft is clearly in control of that
> path).
>
> The text that a user pours into an address bar is just that - text.
> Whether it is ultimately used as a DNS name is up to the transformation
> engines in the browser.  Using Internet Explorer as an example - depending
> on one's "search" settings, the text entered is kneaded umpteem different
> ways to see if it forms a URL that fits a certain Microsoft defined,
> pattern, whether it might appear to be a fully qualified domain name
> (again, ending with one of a series of Microsoft chosen to be recognized
> top level domain names), or is simply a sequence of words connected by
> dots and hence fodder for some kind of search engine, etc etc.
>
> As the result of these transformations, DNS queries that may be emitted as
> the result of something the user typed into the address bar may be quite
> different than the words the user typed.
>
> A lot of people think that any sequence of words connected by dots is a
> domain name.  That's simply not true.
>
> (In fact it is amusing to note that the dot is not a part of DNS at all -
> the DNS protocols do not use a dot as a separator in the packets and can
> actually carry dots (and any other character including blanks) as
> characters in DNS names that are not used to identify host computers.
> The dot that we type is merely part of the human interface.  Well, so much
> for Sun's marketing slogan "We are the dot in dot com". ;-)
>
> I do not like ICANN's proposal to use time at the public meetings of the
> Board of Directors to do these kinds of tutorial.  Not only does it
> transform ICANN into a trade show (already dubbed ICANN-dex by someone in
> the press.)  But, more importantly, it reduces the already far too small
> time available for the public to interact with the Board.
>
> However, it is useful for ICANN to make a survey of non-DNS naming systems
> in order to better clarify the limits of ICANN's bailiwick.
>
>                 --karl--

For archives see:
http://www.interesting-people.org/archives/interesting-people/