Re: Last Call: <draft-farrell-perpass-attack-02.txt> (Pervasive Monitoring is an Attack) to Best Current Practice

[Eliot Lear <lear@xxxxxxxxx>]

Hi Brian,

Speaking as a member of the IAB but certainly not for the IAB, I have to take issue with you on a number of points below:

On 12/13/13 2:54 AM, Brian E Carpenter wrote:

Tom,

On 12/12/2013 23:23, t.p. wrote:

Jari

I am wondering what the role of the IAB is in this.  Statements of
policy such as this I have seen previously from the IAB, as in the
RFC2804 that has just been referenced.  Whereas the IETF produces the
engineering, such as TLS or IPsec, which is rather different in nature.
Does the IAB approve or disapprove of this?  Why isn't it involved?

Like RFC 1984 and RFC 2804, this one is intended to be jointly
issued by the IAB and IESG. I don't see any difference (except for
the artefact that it has to be assigned to one of the streams,
which didn't exist when the two previous RFCs were issued).

Sort of.  The current plan of record was for this to be issued as a BCP.  The document is currently in IETF last call, and then the IESG will decide what to do with it.  The intent is for the IESG to allow the IAB to insert an IAB statement if we wish to have one in there.

And when I look at the IAB website, I am bemused.  The IAB is calling
for papers for a conference on this precise topic, to be held in three
months, by which time you want this I-D to be signed, sealed and
delivered.  

Yes - this is a statment of principle. We can continue to waste time
wordsmithing, or we can just put it out there and save bits.

Well actually there is confusion about this, which is in part why there is a debate.  We've already seen one working group chair expecting the IESG to take actions on documents based on this statement of principle.  And so some care is therefore required.

So that the IAB can wave it at all participants and say

'Discussion over'?  Or what?

The discussion of the principle *was* over in Vancouver.
Workshops, and IETF WGs, have to apply the principle to actual
technology.

I couldn't disagree more.  The discussion over the matter BEGAN in Vancouver.  Nobody – NOBODY – should stop a discussion of principles on this list.  And just to prove the point, you concurred with Dale Worley at least on what one of those discussion points should be.  And to take it even further, and this is only Eliot's opinion, I suspect agreement can be found on the following points:

• Pervasive surveillance represents an attack on the Internet in as much as large amounts of information that is intended to be confidential between sets of individuals is in fact gathered and aggregated by third parties.
• Such a broad scale attack can undermine confidence in the infrastructure, no matter the intent of those collecting the information.
• This is a hard problem.  The very nature of some of the key functions of the Internet allow for pervasive surveillance.  It will NOT be technically feasible to eliminate all aspects of pervasive surveillance.
• Those working on protocol development should take seriously the threat, and conscientiously weigh all of these considerations as they develop technical specifications.
  

The practical ramifications of that last point is that at the very least, WGs should show their work in terms of how they weighed the issues, even if they couldn't solve all of them.

It seems to me that this I-D is an ideal candidate to be presented and
discussed at the conference after which, the IAB can produced a
carefully considered document.

I hope the workshop will be discussing specific technology, or at
least specific technical guidelines, not wordsmithing the general
principle.

Certainly not wordsmithing, but in fact workshops are often for teasing out general principleS (there are a few in play).

Eliot