Re: ISO 3166 recomendations.

["D. Chiodo" <djc@microwave.com>]

On Tue, 24 Dec 1996, Simon Higgs wrote:

> Date: Tue, 24 Dec 1996 02:28:16 -0800
> From: Simon Higgs <simon@higgs.com>
> To: "D. Chiodo" <djc@microwave.com>
> Cc: Kevin Brown <kevinbr@netcomm.ie>, "Rick H. Wesson" <wessorh@ar.com>,
>     iahc-discuss@iahc.org
> Subject: Re: ISO 3166 recomendations.
> 
> At 5:21 PM -0500 12/23/96, D. Chiodo wrote:
> 
> > On Mon, 23 Dec 1996, Simon Higgs wrote:
> >
> > > At 1:20 PM +0300 12/23/96, Kevin Brown wrote:
> > >
> > > You want what? (ROOTFLMAO)
> > >
> > > :-)
> > >
> > > You'd think they'd have figured this out by now.
> >
> > I'm not entirely sure who you are responding to, since the writer of the
> > message that you "replied" to, didnt indicate a "want" for anything.
> >
> 
> Actually, I think you missed 101 newdom. The whole issue over
> governance and accountability of the "." administration has been
> skillfully avoided over and over again. Noone has been able to resolve
> this issue and so the IAHC exists. Technically IANA hold the root, but
> Jon has seen fit to set up the IAHC to obtain input and concensus from
> the community. Having said that, IAHC's focus is very narrow.

Lots of people expect lots of things from IAHC. John Palmer expects them
to put his domains in the root, becuase he feels he has some legitiamte
claim on them. I expect them not to, becuase he does NOT have any
legitimate claim on them. Obviously there are confliciting expectations,
and far more of them than mine and John Palmer's. Nothing the IAHC can do
will make everyone happy.

I will admit, that I do not feel what they have proposed to do in the Dec
19 draft goes quite far enough, and introduces even more confusion instead
of really answering any questions. But I don't really blame them,
considering the lynch mob waiting outside their doors.

But, I think some good things the IAHC could do would include:

Defining (or reemphasizing) what party will administrate the root zone.

Defining connectivity and reliablity specifications for the primary root
zone server, and set a time limit within which the current operator must
bring the server within compliance if it is not. (Since my suggestion for
this would be leased line connectivity, via differing carriers, to at
least 3 of the NAPs with full defaultless routing, a double-redundant
power supply, an alternate server ready to use at all times, and a
bi-daily tape backups (once immediately before root zone updates, once
immediately after, I have a good idea NSI might need to make some
enhancements) 

Defining what will happen to the .COM zone. (My suggestion would be a
phased retirement, and a 5 year moratorium on its recreation)

Defining a process whereby qualified companies and organiations could
apply for registration of a zone within the root (TLD), what those
qualifications are. I would suggest a moderate fee to be placed in a fund
for root server operation, upkeep, and administration (including these
registration procedures) 

Defining what should happen with the "list" of organizations that have
attempted to register and use TLDs in advance of this policy. (My
suggestion would be to permit them to apply and give them preference over
other organizations requesting the same TLD's, but to not relax any of the
requirements, qualifications or fees.)

I would suggestions that the qualifications for registration of a server
for a TLD zone include:

 One primary zone server, configured and ready to operate, which must be
operating on a multihomed network with a defaultless routing
configuration.

 Some combination of secondary servers:
  3 secondary zone servers, none of which is to be on the same network or
backbone as the primary, nor each other.
  2 secondary zone servers, one of which may be on the same backbone or
network as the primary. Both networks must be multihomed to different
backbones, and have a defaultless routing configuration.
  1 secondary zone server, not on the same network as the primary. Both
the primary and the secondary must be on networks with direct connections
to at least two NAPS, and be operating in a defaultless routing
configuration.

All servers must have at least single-redundancy backup power.

A published working email address for the technical administrator of the
primary zone server, that must work even if all servers for the TLD fail.
(IE, it must be in some _other_ domain, and must not forward to or
otherwise rely on the served TLD)

A published working phone number for the technical administrator.

I'm starting to wander. I'm sure IAHC and the net community can form a
reasonable set of qualifications for TLD servers. But yes, IAHC and IANA
need to make a positive step to prevent the existence of multiple root
servers all over the place, which would turn the net into a "spotty, at
best" method of communication. And I think if they define who runs the
root DNS, and a method for appplying for delegation of TLD zones, I think
the "registries" will figure out how to work on their own. 

Everyone is throwing this "registry" buzzword around, and it ISNT the
issue. The issue is what TLD zones will be delegated, who they will be
delegated to, and how. Put the technical details in place, assign the
authority, and the suits will figure out the administrative details for
their organizations. 

> > He did, however, albeit slightly unclearly, make a point that many people
> > such as Alternic have been ignoring. There can be only one "." if the
> > Internet is to continue as the useful and viable means of communication it
> > has begun as. Many dispute this fact, but they are nonetheles wrong.
> >
> 
> You're obviously not familiar with the experimental root zone handed
> out by AlterNIC. It does not fragment the namespace. It just added the
> experimental zones to the existing root. Users either get the IANA root
> or the IANA root with the experimental TLDs. That's not fragmentation.

Yes, Alternic illegitimate "." servers include that data that the
legitimate root server has in its root zone. But what if it changes? The
point is, there can be only one official, in-use, primary root nameserver. 
Perform any experiments you want, but please refrain from claiming they
are anything but experiments, and be sure to keep their use to a closed
envirnment. Alternic and company are not doing this, but are instead
trying to overthrow IANA by trying to dupe as many people into useing
their root server instead of IANA's

> Root64 is fragmentation.

The fact that there are multiple root zones is fragmentation, period. 

> > > > If they set there own policy on DNS, why don't they set there own
> >policy on
> > > > IP space? Either someone controls  . and you have to agree on who
> >get under
> > > >  or not. Yo cannot have two seperate classes of TLDs. No country owns the
> > > > Internet, and no country own the IP space, and why should any one country
> > > > have any policy , and be allowed under . ?
> > > >
> > > > Kevin
> > > >
> > > > At 2:50 +0300 21/12/96, Simon Higgs wrote:
> > > > >At 10:34 AM -0800 12/20/96, Rick H. Wesson wrote:
> > > > >
> > > > >> I believe at the newdom bof the AD's both stated that the IAHC
> > > > >> and the newdom bof should not say *ANYTHING* about management
> > > > >> of the ISO 3166 TLDs.
> > > > >>
> > > > >> Suggestions should not be made about a management area that
> > > > >> no one on this committe has the perview over.
> > > > >>
> > > > >
> > > > >No, I'm actually glad they did this.
> > > > >
> > > > >The recommendations to the ISO-3166 domains are spot on. They are just
> > > > >recommendations tho'. It's up to the individual countries to set their
> > > > >own policy. I would expect most free market countries to adhere to
> > > > >them, and those countries with more restrictive regimes are going to do
> > > > >whatever they want anyway. The point is that there is a consistant
> > > > >reference base for the ISO-3166 countries to follow if they want to.
> > > > >The mistake would be not making these recommendations at all.
> > > > >
> 
> 
> Simon
> 
> --
> If at first you don't succeed, skydiving is not for you.
> 
> 
>