Re: [Ianaplan] PTI or not [was ICG request concerning IANA trademark and iana.org domain name]

[John C Klensin <[e-mail]>]

Eliot,

I've expressed an opinion and the comments below try to explain
parts of it that might be unclear.  I continue to believe that
there is a hidden cost in this WG and the discussion more
generally. We have to assume that the main value of the IETF to
the Internet involves working on and producing technical
specifications.  For any of us who do that work and have not
discovered how to squeeze more than 24 hours into a day,
discussions here impose, In the vocabulary of my economist
friends, a marginal opportunity cost on the time available for
IETF technical work.  In more familiar IETF language, the mere
existence of this WG and its discussions are a denial of service
attack on other work.  Maybe it is worth it, but, unless the WG
Chairs or others have a theory about how this discussion can
turn into a recommendation that will reach CWG in a way that can
do so good, I have real questions about how much time we --at
least those of us who have to postpone or drop other work to
participate-- should be spending on the discussion.

I've heard, and respect, the argument that we have no choice
but, if that is really the case, I hope someone will make the
point to NTIA and ICANN (perhaps at this afternoon's forum) that
this effort is actually expensive and causing damage to the
Internet.    

>From my point of view, and probably obviously, the concern about
time and resource consumption and marginal costs suggests that
the entire community should be trying to resist overly-complex
models and proposals.  But YMMD may well different on that
subject.

More inline below...

--On Sunday, June 21, 2015 09:04 +0200 Eliot Lear
<[e-mail]> wrote:

>...
> On 6/21/15 6:26 AM, Brian E Carpenter wrote:
>> On 21/06/2015 15:54, John C Klensin wrote:
>...

>>> A month or two ago, I didn't see much point in PTI.  I saw
>>> some risks but felt that it was appropriate for the IETF to
>>> be flexible about the PTI idea as long as the (at that time)
>>> missing details didn't create new problems or drag us into
>>> situations that would make our environment significantly more
>>> difficult without adding clear value.  Now we have with this
>>> trademark and domain issue, including the claim of
>>> exclusivity, and a sense that PTI isn't an names-only issue
>>> but something that is necessarily proposed to be imposed on
>>> all three customer groups and the broader community.  
>> Yes, I think that is the only possible interpretation of the
>> proposal.

> The trademark issues have been there all along, by the way.

Of course they have.  But, at least for me, there is a huge
difference between "there has never been an intra-community
problem with the trademarks and their use and there is no reason
to believe that anything about the transition will affect that"
or even "let's use an organization to hold the trademarks that
has a charter and fairly long history of holding resources for
the Internet community to prevent those resources from being
used to, e.g., deny access to them or impose conditions on that
access" and starting to take about transfers to an untested and
unknown organization with an apparent mandate to actively manage
those trademarks with exclusive licenses, etc.".  Could we,
given an adequate supply of time and the right sorts of lawyers,
sort those difference out?  Yes, I think we could.  But see the
above comments about costs and my earlier comments about lack of
compelling justification for the benefits of these changes
sufficient to justify those costs.

> And we just chose not to provide text in our proposal to
> address them becuase someone thought SHMIANA would serve just
> as well.  I don't subscribe to that logic because of our
> existing works, but that was where rough consensus led us.

Again, see above.  From my point of view, we are either going to
reopen that question given apparent new data, we are still
satisfied with that consensus and either don't believe there is
new data or that they are not significant; or maybe we are
wasting time and resources.  The default hypothesis is, of
course, that the earlier consensus stands and gives the WG
leadership adequate information as to how to respond to these
latest proposals on that basis.  If that is, in fact, the plan,
then the next posting I'd like to see is from the chairs and is
either a proposed response or an explanation of why they think
no additional response is needed.

>> If we're going to discuss it here, let me say that it's a
>> very strange logic that leaves the contentious namespace
>> policy issues in ICANN but requires a separate outfit to
>> perform the relatively trivial clerical work at the end.
> 
> Actually that's the GOOD part.  We don't want to be involved
> in the contentious parts and we certainly don't want the IANA
> people to insert themselves in middle of the decision process
> of the contentious parts.  That's for the operational
> communities.

Yes.  I've said elsewhere that, from my point of view, the whole
problem here originates with decisions made during the
development and implementation of the ICANN Affirmation of
Commitments that turned IANA into a policy leverage point.  It
may be too late now to fix that, but, if anything, the broader
community should be striving for ways to reinforce that
separation.  If I thought for a moment that the simple creation
of a more complex corporate structure would help with that, I'd
be in favor of it, but what I see is only a lot of potential for
more complex and less accountable "governance".

>>  I would presume that the work to actually add a new
>> name to the root zone is fairly small and occurs quite rarely.
> 
> Yes it should be relatively easey, and no it happens all the
> time.  That's what the GNSO exists for.

I read Brian's comment as involving the amount of work needed to
add a name once all of the relevant policy decisions had been
made.   I think that is correct and that, relative to  many
other Internet events, even Internet administrative ones or
total IANA actions over the course of a year "quite rarely" is
probably correct. It is also not "what the GNSO exists for" --
relative to decisions about adding new names, the GNSO is only
part of an immensely complex and expensive set of processes.
Unlike the PTI proposal, if ICANN, or its rather larger "names
community" subset has concluded that the current process is
appropriate and/or necessary, I don't think we, even those of us
who believe that some of the results pose a danger to the
Internet, should be second-guessing them in this context.
 
>>  On the
>> other hand, IANA performs quite a lot of work every day for
>> the protocol parameters community, so changing the formal
>> structure around IANA for the sake of occasional root zone
>> updates would be the tail wagging the dog.
> 
> The tail has all the moolah and actually employs the dog, so...

So?  At present, that "tail" does not run IANA or determine
policies that affect the IETF, impose extra work on us, or that
could harm the specification and use of IETF protocols.   If
that is going to change, we need to either insulate ourselves
from the changes or protect ourselves against their effects.
Given, again, the costs of our doing that, I think it is
entirely reasonable for us to ask pointed questions about the
motivation for, and value of, the changes.   If the answer were
"well, we supply all the resources ('have all the moolah and
employ the dog') and we feel like doing it, so you have better
suck it up" then the only rational decision for the IETF is
whether to pull the plug now or wait until later because that
attitude would predict to continue battles.  Fortunately, while
there have been speculative comments on this list about that
kind of attitude, I haven't heard it from a single person who is
associated with the CWG or otherwise taken seriously within
ICANN or the Names community.

>> And afterwards, misbehaviour by ICANN in the approval of new
>> TLDs seems to be a much greater danger than misbehaviour by
>> PTI in editing the root zone file.
> 
> Yes that is why there is an accountability group.

And, given history and track record, I might respond to that
response as if it were a likely and effective solution with a
question about your belief in the tooth fairy.   But I won't,
not because I disagree with Brian but because a division of
responsibility and authority that the IETF has quietly accepted
since at least 1998 (about things over which the IETF itself
_never_ had authority) make that Not Our Problem.  

>>  How does making PTI separately accountable fix that?
> 
> The names community wants the same ability we have- to change
> IFOs when the function is not being performed to their
> satisfaction.  How can we possibly object to them wanting what
> we have?

We can't and shouldn't, as long as their plans to give
themselves that ability doesn't interfere with our ability to
make our own choices.  I note that the separation plans we have
talked about have all been "we pull out and take some or all of
the protocol registries elsewhere", not "we prevent IANA from
functioning unless others adopt our models of how things should
work".  

The PTI proposal was first presented to us as strictly a names
community adjustment to "give them what we have", one that was
not supposed to have any effect on the IETF.   As such, I have a
personal opinion based on a bit of experience with ICANN and
circa a quarter-century of intermittent involvement with IANA
structuring and policy-making that I'd be happy to share (in
more detail than anything I've said on this list) with the CWG
if there were any evidence they they were interested in my
opinion.  But, as far as the IETF is concerned, it really isn't
our problem and we should, if anything, cheer the development of
parallel structures.  However, as this particular tiger's
stripes have become more obvious, it is clear that it is a
proposal that has direct consequences for the IETF-IANA
relationship (I think this trademark issue is a symptom, not the
problem, but an important symptom) and, as such, the objection
isn't their "wanting what we have?" but to their proposing a
solution that could (and probably would) considerably mess up
our relationships, at least unless we invested a lot of
resources in defining and refining new boundaries and models.
That change even makes my questions about motivation and my
doubts about effectiveness relevant because they go to whether
the improvements PTI might make are significant to justify the
IETF investment that would be required to make it work without
negative consequences.

>...
>>> I therefore think the IETF
>>> should go on record as clearly opposed to the PTI concept as
>>> unnecessarily complicated and increasing the risks that a
>>> transition that includes it will turn out to cause serious
>>> problems.  I am mindful of Milton's "PTI or no transition"
>>> prediction but it seems to me that we should not allow real
>>> or imagined issues in the names community to force us into a
>>> situation that we think may be damaging.

>> I think it's reasonable to state that, as long as it's clear
>> that we aren't burning the bridges just yet.

Agreed.  I'd don't want to burn bridges (or houses, see below).
But I do want to make it clear that the current plan, without
more qualification, modifications, or details, is not acceptable
to the IETF because it adds both costs and risks... and that, if
the IETF is asked to incur those costs and/or risks, it is
reasonable for us to question costs to the IETF and the Internet
and, if those costs are justified by the claimed advantages, to
dig into details, motivation, likelihood of success. and whether
there might be less complex and risky alternatives that would
accomplish as much.

> One doesn't start by burning the neighbor's house down when
> one has a different perspective.  One starts by finding a
> means to resolve difference through a conversation by the
> fence or over a cup'a, which should hopefully be happening
> this next week.

An interesting analogy, in part because there is evidence in
some communities of a long tradition of burning nearby buildings
in order to reduce the population and characteristics of people
who would otherwise need to be consulted about proposed new
developments and sometimes to drive down property values to
lower the costs of those developments. 

I would, however, turn your observation around: if I were
embarking on a significant and risky remodeling project on my
house, I would either expect to consult my neighbors in advance
about my plans to be sure that there were no unexpected
side-effects that caused negative impacts for them or to erect a
sufficient firewall (literal variety) to be sure that they
didn't suffer collateral damage from my efforts (indeed, the
laws in many areas require both).  In this instance, there was,
at least to my knowledge, no discussion with the IETF at the
point that the PTI plan was being contemplated.  Instead, it was
sprung on us with reassurances that it wouldn't affect us,
especially negatively (reassurances that I believe were made in
good faith based on the understanding at the time).  It also
came with a very clear "accept it or give up on transition"
comment that was presumably not made on behalf of CWG or at an
official CWG statement but that no one else involved with CWG
has bothered to repudiate.  And then, as the details were filled
in, it because clear that the "no effect" claims were no longer
quite true.

So I see a large fire burning in my neighbor's house and it is
getting close enough to make me nervous and self-protective.

I need to add that "One starts by finding a means to resolve
difference through a conversation [...] which should hopefully
be happening this next week." makes me anxious too.  I'm all in
favor of informal conversations that at least clarify
differences and bring people closer together.  But any
"resolution" cannot possibly be representative of the
conclusions this WG might reach and, in particular, involves
participants who are willing and able to make the investment to
attend ICANN meetings (and more typically supported to attend by
people and organizations with specific interests in specific
kinds of outcomes) -- those populations are not typical even of
the participants in this WG much less the IETF community more
generally.

best,
     john