Where to begin?  Certainly Tony sets the stage with his condescending (and entirely and characteristically inaccurate) “non-organization” reference.  And his reference to the IETF mailing list as being censored is quite simply bizarre.  I can only guess what he might be implying, but I won’t waste my (or your) energy trying.  The reference to the IETF as a DARPA thinktank is amusing; while DARPA might have engendered the IETF, it has never been a direct factor in the IETF’s operation. I was the first independent commercial participant in the IETF, in 1987 and many others quickly followed.  Nor was the IETF ever “adopted”—whatever that means—by the NSF.

Tony’s diatribe about the IETF’s “going after the government hand” would be amusing, if it weren’t so misguided and just plain wrong.  Its silliness on its face is perhaps secondary to its factual error: by 2000, the government was long gone as a primary force in Internet development or operations.  Remember that the Internet went mass-market global by about 1994.

To date, the IETF’s forays into technical policy of the type Tony references are really nothing more than saying that the engineering that it does shall be good engineering.  That’s the essence of the RFC 2804 that Tony cites.  Backdoors make for very, very bad engineering.  And Tony takes exception to a policy of insisting that engineering be done well???

Tony appears to have been hooked by the fishing line of exceptional access, ignoring the profound technical and operational problems it creates that, so far, are insurmountable by the expert community that he seems intent on marginalizing.

All of which is remarkably irrelevant to the current IETF controversy, which concerns venues for IETF meetgins and is trying to grapple with a massive social change happening globally, with respect to the rights and non-rights of a significant portion of any group’s population.  Change like this has happened before.  It never happens quickly or easily and it never happens completely. 

We have previous examples of such changes and none is yet ‘finished’. And during the early stages of the changes—as we are now experiencing—the disparities from one place to another are striking.  Any organization seeking to provide an environment supportive of ‘modern’ views of social rights is faced with challenges of the type type the IETF is wrestling with.  As a small example, ignoring laws on the books carries risks on its face, as does ignoring actual daily practice.

Attending to major social change requires caution and serious consideration. 

Tony’s posting contributes nothing positive to those considerations.

d/

I’m sorry you read the article negatively, Dave.  We obviously characterize the same facts differently - as we have for decades. It is not unexpected.  Perhaps also not unexpectedly, I’ve received multiple notes strongly concurring with the article.  Different circles of friends; different perspectives.  The intent here is to engender a discussion on significant concerns, not determine orthodoxy.  Your involvement is a good start.

You will recall we strongly differed over the wisdom of your setting up an arrangement with the ITU Secretary-General on assuming DNS legal responsibility.  Arguably, my policy concerns expressed there, seem subsequently accurate.

I have had the luxury in participating in a great many bodies over the years, and thread through perhaps a score of them today - rather less so the IETF.  So I have the luxury of looking in largely from the outside, compare it within the larger ecosystem of similar bodies, and look at trends.  I’m simply an analyst raising a note of caution whether you care to recognize it or not as a longtime inhabitant within the IETF. 

You may blow away the adoration at IETF meetings of a thief of classified information, and pursuing his work across multiple IETF initiatives, as inconsequential, and as just “good engineering.”  Many others would disagree.  If actually implemented, some of those steps could cause considerable harm on multiple levels. 

The IETF is on a slippery road of major political activism that has collateral consequences like pondering the appropriateness of venues and their acceptance of social change.  We probably can agree on the humorous admonition that “social change requires caution” - whatever that means. :-)

Trust all remains well after all these years.  Your advocacy in the email history space is great.  ...and, thanks for the introduction to that fabulous nectar, Laksa.

“The controversy is now playing out on censored IETF lists.”
There are no censored lists at the IETF, and most of them are unmoderated, apart from anti-spam precautions and provisions to deal with persistent abusive usage (which, no doubt, CircleID has too). For example, Tony is clearly reading the discussion in question, and he posted to the list a few hours ago.

So now we’ve cleared up that misrepresentation, let’s consider Tony’s wild assertion that the “IETF is on a slippery road of major political activism…”
I don’t think so, and Tony, in the best tradition of polemicists, mixes up two totally different things:
1) Security and privacy, where the IETF has been consistent since at least 1996 in objecting for technical reasons to regulatory measures that damage security and privacy. Nobody in the IETF was really surprised by the Snowden revelations: we all knew such things were possible, but it was interesting to get the gory details of how the 5 Eyes countries were actually doing it.
2) Holding a meeting in a place with laws in place that are unacceptable to LGTBQ participants. That’s a practical question: does such a venue damage our ability to get work done at the meeting.

I see no political activism here. I just see an organisation of geeks trying to do its geek work in an imperfect world.

FYI: I am a former Chair of the Internet Society Board of Trustees, a former Chair of the Internet Architecture Board, a former Chair of the IETF itself, and a current technical participant in the IETF.

The IETF’s following the Snowden agenda and facilitating end-to-end encryption of everything without providing for access to stream information, exposes IETF leadership to considerable penalties. Bonding with Snowden has significant risks for the IETF and its leadership.  The insurance amounts provided by the Internet Society in the early 90s are likely inadequate under the new political-driven agenda.

For example, Cox recently received a $25 million judgement against it for violating the Digital Millennium Copyright Act (DMCA) safe harbor provisions by failing to adequately implement a repeat infringer policy. Consider a litigation hypothetical where one of the well-financed copyright groups brings an action against the IETF leadership for contributory infringement by developing a multi-stream end-to-end encryption protocol that prevents identification of infringers.  The available IETF insurance money might provide a significant incentive.

Consider another hypothetical where somebody suffers major personal and financial harm in a terrorist attack where the non-discovery of the perpetrators resulted from an encryption capability devised by the IETF explicitly undertaken to prevent critical information from being discovered.  The parties suffering the harm or their heirs bring a suit for consequential damages against the IETF leadership.

There are many similar examples that are consequential to political-technical agenda being pursued within the IETF.  The causes of action can also potentially be instituted in any jurisdiction worldwide.  Even if an award doesn’t occur, the litigation costs can be substantial.  Last year, in a ground-breaking case, ETSI was sued for contributory infringement in a standards development case.  Although an agreement was eventually reached, the litigation cost ETSI more than a million dollars.

Tony, it is a little difficult to follow your comments and purported concerns, given that you continue to toss off so many unanchored references, fail to respond to specifics and fail to offer them.  Still, you have nicely provided an excuse to remind you and others that the technical work of the IETF has quite a bit of historical basis, long preceding any current furor.  Possibly relevant to issues you are seeking to touch:

1.  Rather than being in response to anyone’s purported recent ‘agenda’[*], the IETF has always valued end-to-end design, which was best articulated in:

    Saltzer, J.H., Reed, D.P., and Clark, D.D., “End-to-End
    Arguments in System Design,” ACM TOCS, Vol 2, Number 4, November
    1984, pp 277-288.

    http://cyberlaw.stanford.edu/e2e/papers/Saltzer_Clark_Reed_Originale2e.pdf

2.  The technical and operational impediments to ‘exceptional access’ have been amply documented in:

    https://www.cl.cam.ac.uk/~rja14/Papers/doormats.pdf

and elsewhere.  I will remind you that it was written by quite a long list of remarkably accomplished security experts.  So anyone claiming that exceptional access is safely feasible needs to provide a competent design and gain agreement among technical experts.  Absent that, claims that it can be achieved with realistic levels of safety fall into the realm of fantasy, not legal liability.

d/

[*] The reference to Snowden is curious on two counts.  First, he’s never been part of any IETF activity, though there was a skype’d session with him outside of an IETF event.  More significantly, what the heck is the reference to “the Snowden agenda”?  Where is this documented and how is it relevant to IETF work.  Details, please!

I’m having somewhat the same problem in your direction, Dave.  Let’s ascribe this divergence as we have so many times over the decades to different backgrounds, conceptualizations, and weighing what is relevant, and agree to disagree.  As to your two well-known citations, there are obviously significant divergent perspectives on encryption that get played out constantly in many different fora - confirming that the subject matter is indeed political, and that the IETF leadership ascribes to one side of that debate.

As to agendas, RFC 7624 speaks for itself and makes clear it serves as a benchmark for all subsequent IETF work.  The reference to “Snowden revelations” and entire clauses wrapped around his stolen materials also establish the character of the agenda.  It is difficult to spin this other than a precipitous descent into a political rabbit hole.  A less gracious characterization is that it makes the IETF participants potentially criminally or civilly complicit in the ensuing loss of critical capabilities under the laws in most countries concerning financial services, intellectual property, cyber security, law enforcement, and national security.  Most vendors and providers will likely eschew implementing the resulting specifications, or work around them.  As already opined digression into political activism is toxic.

That’s all.

Tony,

“there are obviously significant divergent perspectives on encryption that get played out constantly in many different fora”:  I will ask you to move from your pattern here, of making broad claims that fail to include any specific anchor, and will challenge you to document a view that counters that of Keys Under Doormats and has been formulated by anything close to an equivalent body of technical experts.  Again, absent that, your pronouncements fall in the category of wishful thinking—or political ploy or…well, nothing constructive—and not technical and operational reality.

As for RFC 7624, it’s actually quite easy to ‘spin’ it as something other than going down a political rabbit hole:  for a goal of privacy, pervasive monitoring introduces a new category and degree of threat.  That’s a technical assessment, not a political one. 

As always you are encouraged to counter with substance, though your pattern through this thread doesn’t provide much hope.

d/

On the matter of Keys Under Doormats, the answer is simple.  It is a nice academic exercise similar to many similar ones - almost always motivated by ongoing political developments on which they have their own views.  However, those keys under doormats enable legal obligations to be met with minimal risk of adverse consequences.  Put another way, the comparative benefits of having access to the keys is so divergent that few competent officials responsible for national security have ever weighed the matter differently.  There is a thousand year history of materials over at the National Cryptologic Museum that support that assessment, as well as long histories of international agreements.  One also wonders what some of the Doormat authors would have been doing during the last World War - helping the Third Reich improve their encryption?

As for RFC 7624 - that was undertaken by those whose values are very different than my own.  It is action that seems on its face egregious and immoral; but like you say, it can be spun.  As to countering and hope, the avenues lay in not convincing you, but in the pushback and counter measures of countless bodies other than the IETF, in the business decisions by potential vendors and providers, and in legal systems that will impose penalties on the implementation of the Snowden agenda of RFC 7624.

Tony, that’s impressive.  You wave off the forty years of practical expertise in that document, by the folks who literally invented this field of security and you feel competent to class their work as merely ‘academic’.  Even better is that you also fail to supply any concrete reference to substantive technical work that counters their diligent and detailed effort.  Since it doesn’t exist, that’s not a surprise.

And since RFC 7624 was the product of the usual open IETF processes, where was your substantive input to its development and the input of those agreeing with you?

As for ‘spinning’ that was your term, not mine, and indeed it’s all you seem intent on.  My term is “technical analysis”.  That’s what the RFC provides.

What is impressive is the assertion that the latest article written by a group of folks with a well known political-technical agenda they have been pursuing for years, is cited as the definitive finding on the subject.  Next time you are in the neighborhood, try visiting the David Kahn collection at the National Cryptologic Museum.  Or, get a copy of David’s book, The Codebreakers.

There is no way that Snowden agenda of RFC 7624 can be spun as anything other than a toxic gambit for the IETF that dispenses with its technical analysis stature, and makes it complicit in the grievous adverse consequences.

Now to something more pleasant like some good wine and cheese.  That’s all.

So, Tony, you seem to be resorting to nothing more than tired repetitions of your tired, vague, scatological utterances, with a notable lack of even the smallest effort to anchor your accusatory innuendo in any sort of documented substance.  You keep referring back to the experience of prior decades, and indeed, I’d forgotten just how little substance you bring to debate.  Thanks for the reminder.

Following up on my own idea of solutions being technical and problems being political, I’d like to draw attention to another distinction which has been more or less overlooked in this argument so far: the distinction between avoidable vs unavoidable politics.

If you’re documenting technical solutions to political problems, as is the essential nature of security-related engineering, then defining the problem is a necessary part of the process. In fact, it’s not only necessary, but important that it be done well. Opposing ideologies cannot reach a compromise on this part of the process: the problem statement will be an ideological instrument, acceptable to some, anathema to others. The only way around this issue is to have separate politically-aligned engineering groups, each producing their own problem statements and associated solutions.

If the IETF (or whatever other engineering body) refuses to support more than one standard along these lines (with clearly distinct problem statements), then it is engaging in partisan politics as well as engineering. To say, “we already have a security standard, we don’t want to create confusion” or similar is all very well, but doesn’t make the position non-partisan: exactly one ideology has received official blessing. The ideological element is unavoidable, but accusations of institutional politics can be avoided so long as solutions are judged on their technical merits, and competing solutions are excluded only to the extent that they actually identify the same problem.

I think it’s probably fair to accuse the IETF of engaging in partisan politics with regards to its framing of the security problem. Only an ideological opponent would call it “anarchist”, but it frames government as a potential threat rather than a trusted ally. If you’ll permit an oversimplification, the analysis has always been that privacy is an all-or-nothing proposition, so total privacy is the goal, including privacy from government eyes. Those who argue for authorised back doors have been rebuffed with the claim that this is effectively no privacy—a fig leaf, as it were—and excluded from the process.

Having said that, I think it’s an open question as to whether the IETF ought to be politically aligned in this regard. I think it is so aligned, as Anthony has said, and perhaps it would have been better if it had avoided such alignment in the first place, but here we are, and going back may not be feasible. The problem as I see it now is one of recognising the politics for what it is, and containing it (if not reversing it).

I’ll deal with the “containment” aspect first, even though recognition must come prior to containment, because the containment problem is (I hope) easier to understand. I’ve mentioned cases where a political problem statement is unavoidable, because the problem statement is intrinsically political and absolutely necessary to formulation of a technical solution. In other cases, however, politics can be avoided, and the issues with the Singapore venue are a case in point. This would seem to be a crossroads at which the IETF can choose to contain its political scope, or become further politically aligned, this time along LGBT lines.

It’s one thing to align on issues where you need to agree on a problem statement in order to find a technical solution; quite another to do so where this is not necessary. To do so alienates participants who might have been able to contribute to the problem. And if this unnecessary piece of politics enters the picture, then what next? Arguments will ensue as to which political alignment is the right one, starting with the easily-won battles where few oppose (and simply choose to leave quietly rather than face overwhelming opposition), but ending in the mire of political bickering among nearly-but-not-quite-like-minded folks determined that their precise framing of the world’s social issues is the right one.

A rabbit hole indeed, and note that at no point along this progression does anyone actually recognise that there are any politics in play. It’s all just a “principled” attempt to do the right thing. That brings me back to the other issue: recognition.

I note that in this discussion so far, Dave Crocker has (not unfairly) accused Anthony of “toss[ing] off so many unanchored references, fail[ing] to respond to specifics and fail[ing] to offer them.” On the other hand, Dave has conspicuously failed to acknowledge the presence of politics in the IETF, sticking to the line of it being purely about the engineering. This lack of recognition is a problem because without recognition there can be no containment, as discussed above, and also because it results in a stone-wall attitude of, “this is not politics, it’s engineering, and reality is not negotiable.”

If you’re going to engage in politics, you should bring some diplomacy skills to the table. The IETF rank and file are, generally speaking, very poorly suited to this task. There tends to be a trade-off between innate social skills and innate technical skills, and IETF participants will naturally gravitate towards the latter end of the spectrum. Stone-walling is all very well for a politically powerful entity, but it’s a poor strategy elsewhere: it invites harsh opposition from ideological opponents who are more powerful than you, and they will bring political weapons to the fight, like lawmaking and litigation—things which can and should be avoided by playing one’s cards right.

The problem at this point, however, is one of failing to recognise the problem.

Just two clarifications.  The reference an anarchy has two connotations here.  A certain technical anarchy has always been part of the IETF’s DNA since its early DARPA days - by design. The intent was to generate new ideas as a DARPA funded and led activity, and it was very successful in the face of the rather constrained other internet efforts ongoing at the time in Europe and within the larger industry-government community as OSI.  Anarchy in the form of pure politics by some participants was also threaded through the IETF in later years.  There is copious written material and books on this topic.

The other clarification goes to surveillance capabilities as essential requirements for all kinds of needs beyond just the usual government intelligence acquisition.  It is essential for cyber security, for law enforcement access to evidence, for implementation of the Digital Millennium Copyright Act and its equivalents worldwide, for effective network management by providers, and for implementation of financial securities regulations.  Some of the requirements have existed since the beginning of social structures, and they are not going to diminish.

Excuse me. This is not the IETF that I bought into.

1. IETF has been accepted - in some cases reluctantly - as the prime source of Internet standards. For that, an exclusively technical mandate has been required, and as far as I know, generally provided.

2. IETF is a significant partner in the IANA transition and ICANN accountability. In several respects: PTI, IPR … For that to be sustained, a high degree of stability and predictability is required.

So … please get the IETF act together.

CW

PS: Having participated in IAHC POC, IFWP, ECPOP and the initial 1997-98 ICANN transition, I could enter this thread at some length, but No. NOT in an IETF context.

Hi Christopher,

This is not the IETF that many people bought into.  With this said, a lot has changed over the past 20 years.  It is no longer clear what, if anything, the IETF is the prime source for.  It is one of reasons why it was so easy for Snowden and his followers to hijack the IETF and move forward with significant agendas - as is patent in the cited RFCs and many IETF group threads today.  Try a Google search on RFC 7624 or RFC 7258 at the IETF.ORG site.

Today, the global mobile industry promulgates its own internet standards.  The cable industry promulgates theirs.  Major industry players promulgate theirs.  The rapidly growing NFV-SDN industry which is massively shaping the future infrastructure - promulgates theirs in several non-IETF venues.  So what is left for the IETF?  That is unclear; and going down the political rabbit hole doesn’t help.