DNSO Mailling lists archives


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

RE: [ga] root vs roots - the false but usefull confusion

> From: Dassa [mailto:dassa@dhs.org]
> Sent: Wednesday, May 30, 2001 5:03 AM
> |> -----Original Message-----
> |> From: owner-ga-full@dnso.org On Behalf Of Jefsey Morfin
> |> Sent: Wednesday, May 30, 2001 9:15 PM
> |> To: ga@dnso.org
> |> Subject: [ga] root vs roots - the false but usefull confusion
> <snip>
> |> It is not the root of the iCANN: the iCANN has deliberately
> |> chosen to have its mini-root (authoritative on its TLDs] 
> being only a
> part of
> |> the full and unique root.
> |>
> As ICANN has the major share of the name space the above is a little
> misleading.
> There has been a bit of discussion on ICANN using US law to 
> control other
> roots, mainly in the form of making it illegal for ISP's in 
> the US to use
> them.  This has been touted as not being possible.
> Has anyone given any thought to other legal means ICANN has at its
> disposal?  I would think that ICANN would be able to swing contractual
> arrangements to control access to the legacy root.  Whereby 
> anyone who uses the legacy root is under a contractual obligation to 
> acknowledge it as the authoritive unique root and to not use the legacy
> simultaneously with another root. 

Inclusive root systems don't work this way. What they do is grab a copy of
the legacy root and add to it, their own data. They don't run multiple
roots, yet. They actually run an integrated root. The issue is, what pieces
they integrate into their root. I might point out that most large nets do
this to define their own LOCAL TLD, in a tiered DNS environment. It is not
an uncommon practice. I know, for fact, that SpeakEasy.Net does this and I
have clients that do this (I sold them the system and set it up for them ...
they're quite happy about it too). Those same entities also use HOSTS files
to fix known server constellations, on those same servers, in case the DNS
goes off-line. This has been known to save their butts when the
root-servers.net system goes insane (which it still does, occasionally, and
probably will do again, in the future). This practice also reduces the load,
on the root zone servers, substantially. It is considered a good operational
practice. The only difference between them and the inclusive root operators
are the publication and structure of the zone files. This is also why
"alternate" and "rogue" appellations do not fit them (as much as some would
want it to).

As a side note; they have to do it this way due to deliberate restrictions
in the DNS source code. The iterative query could be modified and extended
into a multiple root zone server search. Such suggestions are vehemently
resisted on namedroppers (even if some of us have implemented it and proven
that it works quite well). Claims that this breaks the Internet is pure FUD.

> They could also knobble the other roots use of 
> the legacy root in the same manner.  Unless there is a contact 
> authorising access to the legacy root, access is unauthorised and illegal.

This last part is definitely of extremely low probability. The legacy root
is published by NSI under government contract. It has been proven, in US
courts, to be a property of the US Government. As such, it is owned by the
US public. As with all documents owned by the US federal government, it is
subject to the US Freedom of Information Act. That means that, unless it is
classified as being in the interests of National Security, the USG MUST make
it available, uncensored and in a timely manner, to any US citizen that asks
for it. The best that they could do is to make it available ONLY to US
citzens. This takes an Executive decree that places the root zone files
under the jurisdiction of the US National Security Act. The National
Security Act is the only exemption, that I know of, to the Freedom of
Information Act. 

This means that it cannot be published on the root zone servers either,
since those are available to non-US citizens. Obviously, that sort of breaks
the Internet. I might also point out that this is something the Internet
community has been staunchly resisting since the days of DARPAnet.

> ISP's would be free to use any root they choose, but to use 
> the legacy root they would be obligated to only use it.

As I said, it's not practical. I hate to sound US-centric here, being an
immigrant myself. But, that's the legal situation, as I know it. The root
zone files and their specific contents are property of the citizens of the
USA and subject to US law and ONLY US law. There is nothing you, or any
non-US citizen, could do about that. Frankly, it would defeat the USG
purpose to do, as you would like. This is why I continually suggest that the
ICANN must first run its own inclusive root zone and zone servers, before
getting uppity with access to the zone files and adding ICANN TLDs. But,
then ICANN becomes "yet another" inclusive root registry. BTW, I originally
registered DNSO.NET for that purpose. If ICANN doesn't do something with it
soon then ROOT-SERVICE.NET shall.

That said, derivative works may be protected by copyright. But, I'll defer
to the IP lawyers on this list with regards to that issue. The derivations
may have to be plainly marked as derivative intellectual property. I don't
think that's been tried before. But again, the IP lawyers on this list
should know.

IANAL - I Am Not A Lawyer. Before taking action on anything I say, you are
encouraged to seek legal advice. 
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

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