DNSO Mailling lists archives


<<< Chronological Index >>>    <<< Thread Index >>>

RE: [ga] collisions in namespace (was gTLD Constituency)

This will be my last post in this thread, as it seems to be going 

It appears that the basic argument is the same one that ICANN 
uses.  They are responsible for only "their root" and all others can 
do as they please.  The fallacy here is due to the fact that there is 
a singular name space which makes the technical argument one of 
fragmentation of that name space.  Once ICANN divides it with a 
collider, it is a done deal.  Then it's a free-for-all.  Technically, 
there cannot be duplicates of any dn at any level in the tree.  Since 
the goal should be that anyone can point to any root and not have 
to wonder which version of a domain they will see, and that ICANN 
is tasked with maintaining that stability, their deliberate entry of a 
collider defeats that task or mandate.

I will agree that one can make a choice to point to any root or TLD 
they choose, but if the goal is a unified name space, ICANN is about 
to defeat that in a big way.  

As for multiple .com's or any other TLD, it doesn't take VC to 
accomplish it.  I takes as much money and effort as it does to set 
up any TLD.  I'd bet that if someone does it (perhaps outside the 
US), it will become populated simply because it's there.  I would go 
a bit further and bet that it would be VERY popular and there would 
be no end to problems in the legal arena among registrants, 
businesses, even to the point of disrupting the global economy.  
Just take an ICANN/new.net approach to it and you have a mess.  

The intent behind the DNS was to make the net singular and unified. 
 If this is defeated by blatant ignorance of the existence of the 
purpose of the DNS, we do have a problem, Houston.  

As for upgrading DNS settings, it is not, IMO misleading to suggest 
that moving from a deprecated legacy version of a rootzone or any 
other outdated root is an upgrade.   ICANN's root is exclusive.  I 
choose to upgrade to an inclusive one (there are more than one).  

Years ago, there were several versions of word processing 
software, but two were the most used.  I upgraded from WordStar 
to WordPerfect because it offered more features and was easier to 
use.  WordStar was still there, of course, for a while anyway.  
However, there was a gradual move toward WordPerfect until 
WordStar just faded away.  They never did get away from their 
legacy base enough to keep the public happy.  It was an upgrade 
to a more useful piece of software, although both were written for 
the same purpose and worked on the same platforms.

The DoC rootzone is maintained in its legacy state because 
ICANN/DoC is reluctant to provide what the public wants - an 
abundance of TLDs.  There is a gradual movement toward other 
roots as a result.  Upgrading to a more modern version of a 
rootzone that offers more to the public is a natural progression.  
The task at hand for those roots, IMO, is to continue to work 
toward a non-colliding situation, regardless of how many TLDs are 
carried.  None should collide in the end.  Carrying a subset is fine.

There may be an upgrade available for BIND users as well - 
MultiBind.  This could be a very interesting turn.  Would you say 
that it would not be an upgrade?  I would most assuredly call it an 
upgrade.  If you choose to take your PC into the service center to 
switch to a larger hard drive, is that an upgrade?  Yes.  It's the 
same platform, larger capacity.

The present situation is exacerbated by the ICANN decision to 
duplicate the .BIZ TLD.  That sent a message to the community 
saying it's okay because their's is a separate name space.  As a 
result, companies like New.net pop up with colliders, making the 
efforts of the other roots to eliminate colliders rather more difficult, 
if not moot.  Name.space did much the same thing by allowing TLDs 
to be randomly added without regard to existing TLDs.  It also uses 
a 4LD (adding .xs2.net to a domain name) to allow resolution for 
those who don't point to specific servers.  It's a totally different 
model from all other roots.  Perhaps that is where all of this will go 
now that ICANN has set that precedent.  I hope not, but time will 

OTOH, if DoC chooses the path that preserves the name space and 
avoids a duplication/collision, then a different message will have 
been sent and the community can continue to work toward that 
goal.  Having the largest market share infers a great responsibility 
to the pubilc.  They can either stabilize or fragment the Internet as 
we know it.

Of course, the simplest solution is for DoC to include all TLDs, as 
the inclusive roots do.  It's not difficult to do.  Then the multiple 
root system would be what it is meant to be - a distributed system 
of roots, carrying the same TLDs or subsets of those TLDs.  It 
would balance the load and provide that much more stability.  And 
ISP in Australia could use a local server instead of one in Asia.  It is 
technically sound.  Based on statements coming from ICANN, I 
doubt we'll see this any time soon.  The special interests will not 
allow it.  



> Comments below...
> > -----Original Message-----
> > From: owner-ga@dnso.org [mailto:owner-ga@dnso.org]On Behalf Of JandL
> > Sent: Tuesday, April 10, 2001 9:49 PM To: ga@dnso.org Subject: RE:
> > [ga] collisions in namespace (was gTLD Constituency)
> >
> >
> > > > -----Original Message-----
> > > > From: JandL [mailto:jandl@jandl.com]
> > > > Sent: Tuesday, April 10, 2001 6:53 PM
> > > > To: ga@dnso.org; jfield@aaaq.com
> > > > Cc: jp@ADNS.NET
> > > > Subject: RE: [ga] gTLD Constituency
> > > >
> > > >
> > > > There is a technical collision in the name space.
> > >
> > > I respectfully disagree.  I can only conclude from your comments
> > > below that it is a marketing collision (or problem), if anything.
> > > Here's why I believe this to be so:
> > >
> > > Anyone that has registered a .biz domain name has been forewarned
> > > that "the new top-level domains require that either you or your
> > > ISP have UPGRADED from the functionally obsolete ICANN Legacy
> > > Namespace to the ORSC INCLUSIVE NAMESPACE in order for you and
> > > your visitors to see the domains you are about to register!".
> > > (http://www.pacificroot.com/register.shtml)
> > >
> > > Of course, I would suggest that the use of the term "UPGRADED" in
> > > the quote above is rather misleading to the consumer (FTC
> > > anyone?),
> >
> > It is an upgrade.  The legacy root is old and lacking.  Too access
> > all the TLDs you do have to upgrade one way or another.
> I disagree with your calling this an upgrade.  Here's why:
> If you have five alternative roots, each containing all the ICANN
> TLDs, but each containing a different set of additional TLDs, which
> alternative root is the upgrade?  Seems to me it's more akin to the
> many different "flavors" of UNIX in the marketplace, Solaris, Linux,
> BSDI, etc.  You really can't call each of these flavors an upgrade to
> the original version of UNIX.  To call the "ORSC INCLUSIVE NAMESPACE"
> an "upgrade" to the "ICANN Legacy Namespace" is simply misleading. 
> And, frankly, it is just this sort of misleading marketing that would
> keep me away from having anything to do with a company that promotes
> their service this way.
> > but
> > > putting that aside for the moment, it seems to me that all the
> > > alternative roots have to do is merely convince all the ISPs, etc.
> > > to point to their root.  And, if all the ISPs, etc. want to do
> > > that, fine.  If half want to point to ICANN's root and the other
> > > half want to point to an alternative root, fine again.
> >
> > That's fine, as long as making that choice doesn't cause a problem.
> > With a collider, it will.
> What exactly is a collider and what is the problem that will be
> created?
> >Marketing the root in this context is not
> > the issue at all.  If it were the choice it is today, the marketing
> > is simply to add to what you get from the ICANN root.  Nothing wrong
> > there.
> >
> > I see this as a
> > > marketing problem for the alternative roots, not a technical one
> > > for the entire Internet.  And, unless someone can convince me that
> > > there is indeed a technical problem here, I believe the term
> > > "collision" is inappropriate in this context.  It conjures up
> > > images of failed transmissions, data packets colliding, etc., none
> > > of which seem to be at stake here.
> >
> > What you have is a fractured internet where that choice must be
> > made.  As I said, up to now, carrying the DoC root intact has been
> > standard.  With a duplicate TLD, it will now have to be a choice
> > between versions of the TLD and duplicate domain names throughout
> > the entire tree.  (however, if an ISP wanted to carry the
> > PacificRoot and still carry the ICANN duplicate of .BIZ, it is easy
> > to do so.  They just could not carry both).  The choice should not
> > have to be made, however, and that is the problem.
> I like choice.  For now I point my DNS to the ICANN root.  But I could
> still be swayed to make another choice at some point in the future if
> there's a compelling reason to do so.
> > Consider a half dozen .coms throughout the world.  Shoot, why not a
> > dozen?  He with the bucks for marketing causes the most confusion.
> > Heck, why not?  ICANN is responsible for only the one root, right?
> > And DoC has authority for only its versions of .com/net/org.  If
> > they can duplicate .BIZ, I guess anyone can duplicate .com -
> > multiple times.  Sheesh.
> Yes, there is nothing to prevent someone from setting up a different
> root with a different version of .com.  And, on the surface, that
> might sound like a good idea.  I mean, heck, the ICANN .com has more
> than 20 million registered names...if someone could get even 10% the
> size, that's still a lot of registrations and still a lot of money.  I
> believe there is only one thing preventing attempts at such a business
> model.  There's no VC money around to fund it.  Why?  Ever hear of a
> snowball's chance in hell?
> > I agree, though, that there is every chance that DoC will defy
> > stability, Congress and everyone else and duplicate the TLD.  When
> > things start getting messy and there is a lot of screaming, it will
> > be too late.
> >
> > >
> > > As for creating problems for the ISP and keeping track of two
> > > different .biz names they might be hosting, well, tough...the ISPs
> > > are just going to have to pick a horse to ride.
> >
> > If it were quite that simple, fine.  Let the ISPs scream bloody
> > murder, I guess.  ( I don't really feel that way)  If I were an ISp
> > and my cost for customer service went through the roof over it, I'd
> > scream loud.  When they start with the lawsuits, it'll get more
> > interesting, I'm sure.  You see, it's not the root choice that they
> > will be fighting.  It's the mass confusion over how to host the
> > domains.  Most domain name holders haven't a clue and many ISPs
> > don't either.
> And to that I say, nobody ever said this was a simple business.  Let
> the market forces drive out the clueless ISPs.
> > Do they want to ride the
> > > one that currently has the most viewers or do they want to ride
> > > the new kid on the block with potential but less viewers.  IMHO,
> > > it's a business decision for the ISPs, etc., plain and simple.  It
> > > might even be likened to the VHS/Betamax dilemma for VCR
> > > manufacturers, tape manufacturers, etc. a couple decades ago.  Did
> > > they want to ride the less used but supposedly more technically
> > > advanced Betamax horse or did they want to ride the more popular
> > > but less technically advanced VHS horse. (Note: My comment on
> > > whether VHS or Betamax was more technically advanced should not be
> > > construed as a comment on the technical capabilities of either the
> > > ICANN root or the alternative roots.)
> > >
> > > Anyway, my suggestion to the alternative roots is to think and do
> > > marketing, marketing, marketing.  And if they don't get busy real
> > > soon, somebody new with deep marketing pockets just might show up
> > > right around the corner, steal all the thunder, and create a third
> > > .biz.  Oops, too late!  They already did!  New.Net, anyone? 
> > > (Okay, okay...so new.net isn't offering .biz...yet!  Sorry to
> > > scare some of you out there.  But there is not a thing in the
> > > world preventing them from doing so if they wanted to.)
> >
> > This is true - today.  After all, New.net did it in abundance with
> > 17 colliding TLDs.  We haven't seen too much trouble yet, but wait
> > until .biz is duplicated in the root with the 90 % market share. 
> > That's when we might see it hit the fan in a big way.  Marketing is
> > important for the Inclusive Name Space, for sure.  However, the
> > collisions will become apparent if DoC does its thing.
> The .biz domain names that are invisible to 99% (or whatever the
> correct percentage happens to be) Internet population will still be
> invisible to 99% of the Internet population after the "DoC does its
> thing".  There will be no technical collisions.  The world will go on.
>  And, for the relatively few .biz domain name registrants that will be
> more or less left holding the bag, well...they were warned (or perhaps
> not given the above misleading marketing statement).
> Regards,
> Jeff
> --
> jeff field
> 925-283-4083
> jfield@aaaq.com

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

<<< Chronological Index >>>    <<< Thread Index >>>