David Kessens <david.kessens@xxxxxxxxx> wrote: > > First of all, I am an Area Director who rarely speaks out in public as > I believe that my role is more of listening, mediating and dealing > with administrative issues that make the IETF work. I commend you for that. > However, there are cases where it is important that the community sees > the perspective of an Area Director. Alas, not all will _want_ to see... :^( > RFC 3883 has proven to me as an Area Director to be a very effective > tool to deal with extremely abusive behavior of a select few. I accept that as your perspective. To others, it appears to be a very divisive tool, amounting to a denial-of-service on the functioning of the group which considers it. > draft-carpenter-rescind-3683-01 will set us back in time as we will > need to have the whole management chain involved again to take actions > against abusive behavior by individuals There's quite a difference between calling a behavior "disruptive" and calling it "abusive". RFC 3683 clearly concerns "abusive" behavior: ] ] Notably, in a small number of cases, a participant has engaged in a ] "denial-of-service" attack to disrupt the consensus-driven process. This creates a fairly high bar to claiming consensus to invoke it. > who have been known to move their behavior from one mailing list to > another and who have clearly shown that they have no interest in > contributing to the IETF. > The PR action has allowed us to deal very efficiently with these rare > but existing cases. It takes a serious effort to get them started, > which is very reasonable considering that it is a very serious action, > but it works very well when it finally is in place. I don't believe there's even rough consensus this has worked "very well" in the cases where it has been started. > If you believe that it is good that the IESG spends more of its time > on management issues regarding this kind individuals, ... which I do not... > draft-carpenter-rescind-3683-01 does exactly what you want. David is getting on my nerves here... :^( 1. We're better off _not_ sorting people into "good" and "bad" categories. 2. It's hard to imagine a _worse_ time-claimer than the 3683-Last-Call process. 3. It's a logical fallacy to claim -- were the premise true -- that we should keep 3683 because some misguided folks want to be rid of it. > The IESG will have to have over and over again the same review and > discussion about the same individual whether her/his posting privileges > need to be suspended. False. Per draft-carpenter-rescind-3683, the proposal for "increasing length" must be requested by the WGC and authorized by the Area Director. The "approval" of the IESG should only consider the details of what is meant by "increasing length". IMHO, this could be a one-time policy statement. If the IESG chooses otherwise, it's merely a question of whether they agree with the AD's policy on the meaning of "increasing length" -- not whether they agree with the WGC's call on whether there has been disruption. > I have experienced as an AD that regular suspensions simply don't work > with certain people as they simply don't modify their behavior, no > matter how many warning they get, how many suspensions are used, or > how long the suspensions are. This statement is quite true. > The PR action is designed for these cases, I tend to agree (though I wouldn't try to speak for the designers). > while draft-carpenter-rescind-3683-01.txt doesn't give us any tool to > deal with this problem. I even tend to agree with this statement! draft-carpenter-rescind-3683, however, _does_ clarify that the power of an AD under RFC 2418 shall not be interpreted away by reading into RFC 3934 a restriction of the length of a suspension an AD may authorize: ] ] This mechanism does not permit WG chairs to suspend an individual's ] posting privileges for a period longer than 30 days regardless of the ] type or severity of the disruptive incident. However, further ] disruptive behavior by the same individual will be considered ] separately and may result in further warnings or suspensions. Other ] methods of mailing list control, including longer suspensions, must ] be carried out in accordance with other IETF-approved procedures. The clarification seems necessary, though I hope not controversial. > I would also like to note that the same inviduals who have been > affected by PR actions, also have been the same individuals who have > abused our openmindedness towards our appeal process and who have > filed one meritless apeal after another, tieing up time and resources > of the IESG (and IAB) that could have been used otherwise. 1. We're better off _not_ sorting people into "good" and "bad" categories. 2. RFC 3683 did nothing to restrict the rights to appeal. 3. It's almost always better to reduce the _need_ to appeal than to restrict the _right_ to appeal. 4. I want to reduce the need to appeal. When the WGC calls consensus that behavior is disruptive, there's no _need_ to appeal that call: if the WG members disagree, they can say so; if the WG members agree, there's nothing the IESG _could_ do to change that. > Not everybody in the community might have noticed this, but the IESG > (and IAB) has spend a ridiculous amount of time in the past year or so > on appeals that had no merit whatsoever that came from the same people > who also happen to have been the subject of PR actions (see for > details: http://www.ietf.org/IESG/Appeals.html). In fact it has come > to a point where at least some of us are extremely frustrated that we > have to dedicate as much time as we have been doing to issues like > this instead of carrying out our IETF mission. I certainly _have_ noticed this. David and I merely disagree on whether this is better fixed by reducing the need for appeals or by calling folks names who file these appeals. > Perversely enough, if we adopt draft-carpenter-rescind-3683-01, we > will also need to make much more often decisions to suspend people's > posting privileges opening us up to even more merit less appeals. I suppose David believes this. :^( I believe Brian has been very careful in drafting the language of draft-carpenter-rescind-3683 so as to remove any need for the IESG to discuss the behavior leading to suspensions. I believe he succeeded. I believe that is what most of us want. I do not believe, frankly, that we can wordsmith rescind-3683 to remove the possibility that someone will try to make future IESGs consider that. I'm frankly afraid that we'll get into ratholes trying. I ask David to try to let go of placing the blame long enough to consider which folks are best equipped to judge whether behavior is disruptive, and which folks are best equipped to set policies on the length of suspensions. > The PR action is actually a mechanism that is well documented and that > fits much better in our tradition of openness, as the Last Call makes > such a difficult decision a community decision as opposed to a large > number of suspensions that can be decided within the IESG which in > effect might equate to being a permanent suspension but without > community input. Frankly, I'm not worried about that possibility: there is no such thing as "without community input" -- there are too many opportunities for the "community" to "input" to the IESG members. The particular "community input" method of RFC 3683 is really poorly suited to this problem space: we simply _shouldn't_ be looking for "IETF consensus" on whether a particular person is "abusive". IETF consensus is too hard to reach, and is inappropriate to the question of how an individual WG can function most effectively. > In fact, the PR action tool could have been even more effective if the > sergeant-at-arms of this list would have taken more decisive action by <snip> I wish to specifcally endorse the actions of our sergeant-at-arms. > draft-carpenter-rescind-3683-01 is also mingling two issues > together that should be dealt with separately: > > - it clarifies that longer suspensions are allowed > - it rescinds RFC 3683 The mingling, IMHO, is appropriate. I do not believe David has made a case for either of these without the other. > While I agree that longer suspensions should be allowed, I don't want > to give up the PR action for the most extreme situations. They are > both part of a continuum of options that are all needed, depending on > the situation. David is certainly entitled to write his own I-D for this result: but he has said himself that the IESG shouldn't be trusted with unlimited length suspensions absent "community input". I don't believe there is anything close to IETF consensus on how to do this, so David needs to try to write that himself. Myself, I quite agree with Brian that ] ] BCP 83 [RFC3683] has been found troublesome and contentious in practice. > To make matters worse, this document puts non working group IETF lists > in an even worse spot as the draft somehow decides not to deal with > exactly the same problem for non-working group mailing lists: > > 'Management of non-working group mailing lists is not currently > covered by this BCP, but is covered by relevant IESG Statements.' You lost me, David: I don't see how this "makes matters worse". We _can't_ fix all problems in a single stroke; and insisting on tackling too many issues at once begins to look like merely placing obstacles... > To summarize, I hate the PR action tool as much as anybody else. It is > really sad that we had to use this tool and I am not proud that I > personally had to be involved with one as closely as I was. But there > really is no other alternative for those rare cases where somebody > simply doesn't have any interest in contributing to the IETF, but > instead seems intent on disrupting our work and chasing productive > people away. I am sorry that David has trouble seeing alternatives. -- John Leslie <john@xxxxxxx> _______________________________________________ Ietf@xxxxxxxx https://www1.ietf.org/mailman/listinfo/ietf