t byfield on Thu, 13 Feb 97 06:21 MET

[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

Re: nettime: [Fwd: rewired Zeit- name.space]

At 3:30 PM -0500 on 2/12/97, Gordon Cook wrote:

> Gongrip is absolutely correct <...>

	A few remarks:

	(1) It's silly to speak of "conspiracies" in this context. Protocols
whose functionality relies precisely upon agreed-upon convention are the
very definition of "conspiratorial" practice--because there is *no*
alternative. All this other stuff about SAIC being "CIAs" backward etc.,
etc. is at best irrelevant. [I hope to post an elaboration of this point to
nettime soon.]

	(2) When Paul first proposed "panet" (now "name.space"), he described
it in aggressively anticorporate rhetoric: as those who were present at the
Paradiso impromptu may remember, the project would, he said, be a blow
against "the corporate agenda." It has since mutated into a legal
corporation and justified with "free market" rhetoric.

	(3) name.space's modus operandi--"first come, first served" and
respect for trademarks and copyrights--differs in no substantial way from
the NIC's present policies.

	(4) There are now a number of organizations engaged in distributing
"TLD"s (top-level domains): I turned up agn.net, Root64.net, MCS.net, and
vrx.net in a few minutes of searching. And, of course, there's alternic. As
Rop and many other's criticisms have implied, the failure of these various
upstarts to cooperate could produce a very useless chaos; but their
cooperation will only reproduce the problem of the NIC's hegemony.

	(5) I'm amazed at how sloppily people are thinking  and discussing
this issue. There's no easy answer, but that's not because it's very
complicated; rather, the difficulty is structural. And that's Structural
with an uppercase "s".

	The problem, if indeed there is one (impatience can be problematic,
too), isn't the NIC per se but, rather, *territoriality* in practice and
*scarcity* in theory. Example: My name, Ted Byfield, if fairly unusual, but
I was named after an uncle--and if he were to claim the "ted.byfield"
domain with one of these new TLD services, I wouldn't be able to do so too.
So the universe of possible names has expanded but the theoretical
limitations remain the same.
	This fact can be interpreted in different ways. If you accept the
analogy between a universe of possible names and food (a bogus analogy,
imo), you could argue--I wouldn't--that expanding food production enough to
feed everyone doesn't really matter because they *still* get hungry every
day before supper. Alternatively, consider the analogy to *money* (which
has certainly figured in the genesis of name.space). Imagine a situation in
which anyone and everyone could mint as much as they wanted and could name
and design their own currency and denominations: the vast, vast majority of
the stuff *won't work*. Anyone with two neurons to rub together would
accept only that money which they could reasonably expect others to
accept--since the essence of money, after all, is its fungible quality.
Thus, the salient factor in this case might be seen as *reputations*: Do I
know you? Do I know that others know you? DO I believe that they trust you
enough to accept what you guarantee? (This subject has received an enormous
amount of attention in cryptographic circles; anyone interested should hunt
down an old [pre-1996] archive of the cypherpunks mailing list--and good
luck to you slogging through it.)
	The money analogy is far more salient, but the most appropriate rubric
for understanding DNS is, of course, *language* in general--and
translation, specifically. Obviously, there is much room in "language" for
people to use it idiosyncratically in specific situations while statist
institutions such as the EETS in the UK and the Academie Francaise to
define, delimit, and promulgate national languages on an "official" level.
But things become more complex as the *register* of any given utterance
becomes more system-oriented. It's fine to question or challenge some
institution's "exclusive right" to assign names to a numerical universe;
but that universe needs to *work* while they sort out their differences.
	Here's yet another analogy--one that is oddly close to the name.space
situation. Who has the "right" to assign linguistic names to stars? How
useful would it be for ships or planes to navigate according to charts
drawn up by anyone who felt like "registering" their own star names and
imaginary constellations with anyone who felt like setting up such a
registry? What if one ship is sailing absolutely nowhere because its
navigator has an "art project" map? Or what if two ships are about to crash
into each other because they're arguing over the names and configurations
of stars? As anyone who has put a network together knows very well, these
specific analogies are very much to the point.
	The problem we're dealing with in the context of a TCP/IP
infrastructure is, above all, one of clarity and consensus. Is anyone
fighting over who has the "right" to assign port numbers? Why not? Why
should SMTP be on port 25 or FTP be on 21/25? Or finger on 79 or UUCP on
117? What if everyone decided that the IETF cold go to hell and they'd just
assign their port numbers according to their own tastes? For that matter,
who the hell has the right to decide that web protocol should be designated
"HTTP"? I think I'll call it "TED". And I just decided that FTP should be
called "PIT" and SMTP will be called "GEERT"-- and I'm taking applications
to rename all the other protocols.
	Granted, there are huge differences between port numbers and DNS; but
the only reason that either system works at all is because the procedures
and protocols have been carefully vetted and promulgated.
	The fact of the matter is that there's no shortage of domain names.
There may be a shortage of "vanity" domain names, or of easily memorable
domain names; and there may be a question as to whether the main TLDs are
capable of absorbing the rapid growth of the net; but these issues by no
means require that reform efforts and challeneges to the traditional means
for defining new policies should be entrepreneurial, or should justify
themselves either by rejecting corporations or by appeals to "the market."
	By the same token, Paul's efforts have been far from futile. They have
definitely helped to get people thinking in structural ways, and to address
some basic issues. But, in my opinion, the *fundamental* issues have been
overlooked. Expanding the universe of possible names is a bit like mining:
you may extract more of what you want from the land, but at what price?


P.S. A company based in Edmonton, Alberta, Canada, has started to develop
public-access net kiosks. It's name? Nettime. Fireworks at 11.

*  distributed via nettime-l : no commercial use without permission
*  <nettime> is a closed moderated mailinglist for net criticism,
*  collaborative text filtering and cultural politics of the nets
*  more info: majordomo@is.in-berlin.de and "info nettime" in the msg body
*  URL: http://www.desk.nl/~nettime/  contact: nettime-owner@is.in-berlin.de