Kudos to Adrian and Pete for working on this and producing a thoughtful 
proposal.


I'm separating my comments, and will limit this initial posting to a few 
basic issues.  I'll post pickier quibbles later...



Summary:

      Although reasonable by many measures, the proposal needs more 
complete specification of the standing Ombud function.  It also needs a 
different appeals body.



Background:

      I believe the primary goal, in the IETF's having and processing 
harassment complaints, is to ensure that the IETF is legitimately 
inviting to the widest possible range of participants, and hence to 
encourage their active participation.

      Groups with more diversity produce better work.  Groups that 
suffer harassment of participants inherently reduce real diversity.

      In spite of the IETF's long history of open participation, the 
fact is that daily IETF life often includes explicit and implicit 
tolerance of various forms of harassment.

      The new anti-harassment policy sets quite a good foundation for 
changing this.


However as challenging as it is to produce a policy statement, it is far 
easier than figuring out how to produce a practical implementation of 
the policy...



The Details:

This topic is quite foreign to the experience, skill set and even 
perspective of most IETF participants, touching delicate and painful 
legal and personnel management issues.  Anyone put in a position of 
handling a harassment report needs training, coupled with expert advice. 
  Raw intelligence and good intentions are not enough.

(On the other hand, the principal actors in a harassment complaint are 
not employees of the IETF.  So there might be some additional degrees of 
freedom in the way that IETF-related complaints are handled, compared 
with what a regular company has to do.)


0. Defining Harassment

I think the IESG Statement provides extended text defining and 
describing harassment that is quite reasonable, in tone and scope.  The 
text should be incorporated into an IETF-approved document, which is 
what I'm hoping Adrian and Pete's draft becomes.


1. Ombud Formation

I believe the most critical bit of core work for handling a report of 
harassment is to already have a well-defined and stable Ombud function, 
with good community understanding of the function and approval for it. 
The proposal defers the substance of this issue to the IETF Chair. 
While the Chair is probably the right function for overseeing formation 
of the Ombud, the framework for the office should not be left to the 
vagaries of whoever happens to be Chair and whatever they happen to 
think right at one moment or another.

So the Ombud function needs to be formally defined and according to 
community approval, not merely community review.  Hence the requirements 
and details for the function need separate and stable specification, 
with community support around it.  Then the IETF Chair can reasonably be 
asked to implement it.

It is tempting to suggest assigning the Ombud function to a senior human 
resources professional.

However they will not have the experience with the IETF culture and are, 
instead, likely to apply a classic corporate model for responding to 
harassment complaints, which sometimes includes worrying more about 
protecting the company than about ensuring appropriate handling of 
actual abuse.

So it is better to rest the responsibility for Ombud work on the 
shoulders of experienced IETF volunteers.


2. Appeal

It's reasonable to consider the IESG or IAB as an appeals review board, 
since they already have some IETF appeals responsibilities.  However 
harassment reports call on completely different expertise and completely 
different dynamics.  They also require far greater confidentiality and 
sensitivity than is reasonable to demand from the IESG or IAB.


So, here's a modified proposal...


1. Setup

      a. Formulate a standing Complaints Committee -- called the Ombud 
-- of 3 IETF volunteers.  This is small enough to make efficient and 
coherent workings workings more likely, along with maintain a reasonable 
degree of confidentiality.

      b. Give them the relevant professional manager's training in 
handling harassment complaints.

      c. Recruit a professional HR harassment adviser and a 
harassment-savvy attorney, to act as advisers.


2. Processing:

      a.  On receiving a report, the committee works with the Reporter, 
Target and Respondent to formulate a sanitized summary of the complaint 
and 'defense'.  This will be used if there is an appeal.  It also will 
serve to get the involved parties to focus on the essential details. The 
wordsmithing goal of the sanitization is to produce a version that can 
be shown outside of this initial group.  Possibly only an appeals 
committee but possibly even public.

      b.  The Ombud works with the principals to assess the claimed 
offense and formulate responses they deem appropriate.

      c.  The Ombud modifies the sanitized writeup to include a 
sanitized summary of the decision.

      d.  If any of the principals wishes, they can appeal the Ombud's 
decision. Appeal is to the ISOC Board of Trustees.



3. Appeals:

      For a topic like this, I think one level of appeal is sufficient 
and that the sensitive personnel and legal aspects of this topic warrant 
restricting who is involved, including choosing a venue that already has 
legal strictures in place.  Hence the ISOC Board, rather than the IESG.



4. Addendum:

I'll offer one more item I think basic to discussion of this topic.  For 
now I'm not suggesting specific wording changes, but do want to 
circulate the key point, because I think it has a profound effect upon 
one's thinking about harassment.

We typically think of harassment as being a problem only for the 
Respondent and the Target.  However there is a third, critical 
participant:  The rest of us.

I'll offer a precept that we need to keep in mind:

      Harassment behavior does not merely create a hostile work
      environment for the target(s).

      It creates it for everyone.

To the extent that participants watch others get treated badly and do 
not then see the misbehavior get corrected, there is chilling effect 
across the IETF.


Again, thanks to you both for working on this.

d/
-- 
Dave Crocker
Brandenburg InternetWorking
bbiw.net