ICANN/DNSO
DNSO Mailling lists archives

[council]


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

[council] IAB Statement on IDN


Bruce and fellow Council members:

I have, as yet, received no response to email of last week asking for 
clarification on how ICANN will proceed on the IAB statement regarding 
Verisign's IDN proposal (copied below). I should therefore like to 
request that the matter be placed on the agenda for the 20th.

I wish to stress that I have no wish to debate the merits of the IAB 
statement and that, I believe, ICANN has acted entirely properly and 
within its mandate in consulting the IAB on a technical matter that 
pertains to the stability of the Internet.  What I would ask us to 
discuss, and invite resolution from the ICANN Board, is what happens 
_next._  The IAB statement was in that middle ground where conflicting 
assesments of risk and guiding policy are most conflicted and where a 
clear standard on regulation is desperately needed.  It will not serve 
users for Verisign to wodely deploy, only to be required to shut down 
its IDN system becuase it induces instability in the DNS.  At the same 
time, some risk must be tolerable or we devolve to total statis, which 
is anathema to the dynamic nature of the Internet.

Harold Feld

_______________________________________________
Stuart, Vint and Louis (I have cc:ed the Names Council list, as I expect 
this to be a matter of some interest):

I am writing to ask what follows from yesterday's IAB Response to 
questions about Verisign's IDN efforts.

Let me begin by stressing that it was absolutely proper for the BoD to 
consult the IAB.  This is an issue of technical stability and 
interoperability which lies at the heart of ICANN's purview.  And, if it 
appeared that Verisign's IDN project would cause the DNS to crash and 
burn then action under either the Cooperative Agreement or Verisign's 
registry contract would certainly have been appropriate.

That said, it appears that the IAB did not reach that conclusion.  As a 
lay person, I am not entirely sure what the IAB precisely did say.  They 
clearly had some issues with Verisign's IDN, and in particular appeared 
to favor a uniform approach among gTLDs arrived at via the traditional 
IETF standards.

We now come to the perilous intersection of technical stability and 
policy.  It appears that Verisign's IDN does not create a crisis.  But 
the IAB has issues.  How do we (ICANN) proceed?

There is, of course, a straightforward legal question which lies in 
Louis' baliwick.  What is ICANN legally required to do, and what is it 
empowered to do?  But there is a broader question of what ICANN _should_ 
do.  Arguably, having determined that no crisis exists, ICANN should 
leave well enough alone.  Alternatively, if the IAB assesment indicates 
that there is a potential for real disaster, some prophylactic steps 
might be justified.  At the very least, further discussion is warranted 
and some means of monitoring so as to avert a crisis before it occurs 
would seem desirable.

But if the crux of the IAB objection is that Verisign has moved forward 
without the blessing of the IETF, or that there is some potential for 
consumer confusion, or that all IDN should be standardized for reasons 
other than interoperability, then, it seems to me, a very different 
approach is warranted.

As I have said before, I am very leary of ICANN becoming an industry 
regulatory body.  I have some experience of both public utility 
regulation and media regulation in the United States.  They are not 
pretty, but such intrusive regulation is sometimes necessary.  But the 
entire purpose of ICANN is to spare the Internet from such a need. While 
some have proposed treating gTLD registries as public utilities or 
public assets, this has never been sanctioned as ICANN's mandate. 
Rather, ICANN exists to maintain sufficient stability to allow the DNS 
to both function and continue to evolve.

I would therefore personally urge a course that permits the greatest 
autonomy for gTLDs consistent with basic principles of interoperability, 
technical stability, and fairness to all registries.  Verisign should 
not, of course, enjoy a greater freedom to innovate than other TLDs. But 
innovation in registry-level services is, it seems to me, a positive 
good.  Indeed, one of the criteria in the .org re-delegation was whether 
the new registry proposed innovative services.

In particular, I hope that nothing is done which would appear to make 
conformance to IETF standards mandatory, or which would limit innovation 
  until the IETF has formalized a standard or process.  The great 
strength of the IETF (again, from an outsider's perspective) has been 
its voluntary nature.  "Rough consensus and running code" has fostered 
considerable innovation precisely because it imposed no orthodoxy. 
Anyone with a different approach to a problem was free to try it, and 
this competition of ideas has served the Internet exceedingly well. 
While enough of us rely on the DNS that experimentation cannot be 
allowed to place the Internet at-large at risk, I believe the Internet 
-- and the IETF and IAB -- would be disserved by making IETF standards 
mandatory.

I wish I had a definite suggestion to offer on how to proceed, but I 
don't.  I can only say that ICANN has behaved prudently so far, and I 
hope it will continue to do so.  But we must not equate prudence with 
stasis, nor equate caution with the elimination of all risk.

Harold Feld
GNSO Rep.
NCC





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