| TOC |
|
This Internet-Draft is submitted to IETF in full conformance with the provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that other groups may also distribute working documents as Internet-Drafts.
Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as “work in progress.”
The list of current Internet-Drafts can be accessed at http://www.ietf.org/ietf/1id-abstracts.txt.
The list of Internet-Draft Shadow Directories can be accessed at http://www.ietf.org/shadow.html.
This Internet-Draft will expire on October 26, 2009.
Copyright (c) 2009 IETF Trust and the persons identified as the document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents in effect on the date of publication of this document (http://trustee.ietf.org/license-info). Please review these documents carefully, as they describe your rights and restrictions with respect to this document.
This document reports on the work of the 2008-2009 IETF nominating committee (NomCom). This draft summarizes the process steps that were used this year, and the work that was done. This is followed by a discussion of process issues which caused difficulties for the committee, but which require community agreement to before changes can be made. Finally, there are some observations about things which can help future committees, and which may help the community to understand.
1.
Introduction
2.
Getting Started
2.1.
Randomness and Voting Member Selection
2.2.
Selection
2.3.
Liaisons and Confirming Bodies
2.4.
Tools
3.
Organization, Scheduling, and Planning
3.1.
Job Descriptions
3.2.
Questionnaire
4.
Collection of Feedback
5.
Candidate Selection Process
6.
Issues
6.1.
Volunteer exclusion
6.2.
Liaison participation
6.3.
Nominee List Confidentiality
6.4.
The IAB Roles
6.5.
Multiple Leaders from a Company
6.6.
ADs and chairs
7.
IANA Considerations
8.
Security Considerations
9.
References
9.1.
Normative References
9.2.
Informative References
§
Author's Address
| TOC |
This document provides a set of observations from the 2008/2009 nominating committee (NomCom) chair to the community. These are provided in an Internet-Draft, to complement the report provided by the chair at the March 2009 plenary. This is not a formal report from the committee, and this report represents only one of many possible views of the process.
In discussing the work of the NomCom, this document often refers to the committees judgment of the wishes of the community. (Or other similar words.) The committee is tasked by the RFCs with making such judgments, and acting upon them. As noted, in some places it is even expected to report those judgments as part of confirmation processes. However, these judgments should not be confused with formal determinations of rough consensus of the community, particularly as might apply to changing procedures. It is not the NomCom's purview, nor their right, nor is it practical with the processes that are used, to make that sort of determination. For any process changes to actually be made, separate determination of the actual rough consensus of the community is needed.
| TOC |
The 2008-2009 IETF Nominating committee, like all nominating committees for this community constituted since June 2004, was appointed and operated according to the rules defined in [RFC3777] (Galvin, J., “IAB and IESG Selection, Confirmation, and Recall Process: Operation of the Nominating and Recall Committees,” June 2004.). Lynn St. Amour (ISOC President and CEO) announced my appointment on July 12, 2008.
| TOC |
I sent out the first call for volunteers to serve on the nominating committee on July 15th, with the list of positions to be filled. Several additional calls for volunteers were made, including a presentation at the Thursday Plenary at the Dublin IETF meeting.
In my announcement send on August 1, 2008, I included a planned time line, and the random seeds to be used in conjunction with the procedure described in [RFC3797] (Eastlake, D., “Publicly Verifiable Nominations Committee (NomCom) Random Selection,” June 2004.). The seeds were the same as those used by Lakshminath Dondeti the previous year:
| TOC |
The NomCom volunteer list was announced, with affiliations, for challenge on August 18. No challenges were received, and the list was re-announced (minus the IAOC removal) on August 25. The random selection URL was updated, as one of the announced URLs did not work. On August 30, the results of the random selection, along with the affiliation of the selected individuals, and all of the seed information was announced, starting a challenge period. On September 8 the NomCom was seated and I began working with the members to get the actual selection process in place. The selected volunteers with their announced affiliation, as they were announced by email, were:
Wasserman, Margaret; of ThingMagic moving to Sandstorm
Isomaki, Markus; of Nokia
Hoeneisen, Bernie; of Swisscom
White, Russ; of Cisco
Lepinsiki, Matthew; of BBN Technologies
Hanna, Stephen; of Juniper Networks
Tschofenig, Hannes; of Nokia Siemens Networks
Aboba, Bernard; of Microsoft
Livingood, Jason; of Comcast
Hartman, Sam; of Painless Security
The liaisons were Lakshminath Dondeti of Qualcomm as the previous year's chair, Dave Oran of Cisco as IAB Liaison, Ross Callon of Juniper as IESG liaison, and Bert Wijnen as ISOC Board of Trustees liaisons.
| TOC |
I met with the IAB chair and liaison, and exchanged email with the ISOC liaison, to work out some of the processes. The two topics were what information would be provided for confirmation, and how confirmation issues would be handled.
Even if it appears that everything is simple and clean, such conversations between the NomCom chair and the confirming bodies in advance of the work are a very good idea.
It was agreed with regard to confirmation information that in addition to the descriptions of the candidates and the explanation of why they are being nominated, the nominating committee would provide most of the questionnaire to the confirming body. The exact method for defining what would be kept confidential to the NomCom would be worked out once the NomCom was seated.
With the IAB, it was also agreed that while the IAB would approve or reject an entire slate, they would provide information to the nominating committee with as much clarity as possible as to what the problems were, if there were problems.
| TOC |
One of the most important aspects in practice for the NomCom, and the NomCom chair, is the tools setup. Henrik Levkowetz built and maintains an extremely useful tool suite for the nominating committee. Henrik does not participate in the nominating committee activities. However, nominating committee chairs may wish to ask him to officially serve as an adviser to the committee. This allows him to see information needed to diagnose problems with the tools.
The tools need to be configured properly each year. The first piece of configuration required is a public / private key pair created by the NomCom chair. The public key will be used by the tools. The chair must arrange to deliver the private key to the NomCom members. The email lists are an interesting case. The usual practice has been, and is recommended to be, to use a two part list. The public list is maintained by the secretariat, without an archive. It copies all received email to all members of the NomCom, and to a private list on the tools site. The private list is indexed and archived, and stored encrypted with the years public key. Via the tools, members can get at the full indexed, processed, email information, by providing the private key the NomCom chair gives them.
The tools assist in keeping track of nominations, acceptances (or turn-downs), questionnaires, and feedback. They are extremely useful.
| TOC |
Starting on September 8, 2008 the NomCom organized itself, scheduled conference calls, and began getting its work done.
While I issued the initial calls for volunteers, the NomCom needed to get several items sorted out as quickly as possible. Specifically, the job descriptions needed to be understood, and the questionnaires needed to be defined.
Once those two items were clarified, the NomCom moved on to working out the interview process. Interviews were conducted with a number of people. Some interviews were conducted to understand the needs of bodies or slots to which the NomCom was responsible for making appointments. Some interviews were conducted to better understand existing people, issues, or situations. And interviews were conducted to get opinions on nominees under consideration (within the constraints of the confidentiality rules.) Many people from the leadership bodies were interviewed. Not all the people on any body, nor all the incumbents, were interviewed. Interviews were scheduled based on either the committees perception that incremental information could be obtained usefully, or because an individual requested an interview. While it has the potential to produce an excess of interviews, the nominating committee was able to interview all individuals who requested time from us. Most interviews were conducted by two volunteer members of the committee. A few of the interviews were conducted by larger numbers of people due to the importance of the topic and the potential information. One of the chairs jobs became making sure that folks who had agreed to perform interviews actually did them. Not due to lack of willingness, but because many of the committee members were quite busy.
There were many teleconferences. Initial calls were every two weeks, and then during the busy part of the cycle they were as often as every 6 days. This year Cisco and Microsoft provided the conference call facilities, for which I thank them. We tried a free conference call facility, but it had very bad international access.
| TOC |
The procedures call for the various bodies (IESG, IAB, and IAOC) to provide job description for the jobs that need to be filled. However, the nominating committee is responsible for deciding what qualifications and skills to actually look for, based on the provided job descriptions and the community feedback.
This leads to an interesting question. If the nominating committee decides that somewhat different qualifications are needed than were asked for, what should it do? At the very least, it is required to provide the list of what it actually looked for to the confirming body. It would seem helpful if, when there is a difference, the nominating committee could tell the community what it really wants. We could not find a good way to do this. As there was no significant initial difference (minor issues were discussed with liaisons), the nominating committee simply published the provided job descriptions. Later community input was used to complement those descriptions.
| TOC |
There were two issues the NomCom needed to work out for the questionnaires. First was what questions to ask. This sounds easy, but is actually a good opportunity for the NomCom to start thinking about what matters to the volunteers and how they will gather information. After significant exchanges, the content of the questionnaires was worked out.
It probably was unfortunate that I had posted the previous years questionnaires on the web site, as this confused some people about what they were supposed to do.
In conjunction with this, the nominating committee needed to work out, and coordinate with the confirming bodies, what information would be provided from the questionnaires in the confirmation process. Two specific decisions were made, and worked well.
The questionnaires were explicitly and clearly labeled as to what information would be provided to the confirming bodies, and what would be kept confidential to the nominating committee itself. It was concluded based on prior years that clear communication about the expectation that most of the information would be shared with the confirming body was necessary.
The nominating committee debated what was the best way to ask for private information. The decision was that a single question at the end of each questionnaire would ask for further input not to be shared with the confirming bodies. This allowed folks to tell the NomCom what they thought was important while clearly allowing the NomCom to share enough information with the confirming body to help them perform their job.
| TOC |
The nominating committee can not evaluate the nominees on its own. The 10 volunteers, even with the help of the liaisons, simply do not know all the people, or all the relevant information even about the people they individually know. Community feedback is an essential part of the process.
In order to get this feedback, the nominating committee reviews the nominees, and then sends out requests for feedback. These requests must be done in a fashion that complies with the confidentiality rules in place. The practice that has developed is to send the request to a list of people who may have useful information. The list includes not just people under serious consideration, but also some number of folks who are not being considered, typically because they are not willing or able to serve. This padding is intended to provide some confidentiality to the actual process.
In some cases, because the list of nominees was very long, the committee also made a first pass at removing people from the list before soliciting feedback. This is useful to help the community focus where it will do the most good. But it has the drawback of requiring certain preliminary decisions (which can be changed later) before there really is sufficient information.
It would probably have been useful if I could have found a good way to tell the community what constitutes helpful or unhelpful feedback. Helpful feedback is specific and clear. "Person X did action Y which had benefit / drawback Z." This can be coupled with indications as to whether the commenter thinks this was isolated, or a pattern (for good or ill.) General descriptions of patterns of positive or negative behavior are helpful, but nowhere near as helpful as specifics. General comments like "A would make a good IAB member" or "how could you possibly consider B for that IESG slot?" are actually relatively unhelpful. The NomCom process is not voting, and the committee does not count support. While it means something if several vague positive notes are received, and no negative ones, it does not actually help the committee evaluate whether this person would make a better or worse candidate than some other nominee. Another common kind of response that is less helpful on its own is simple ranked lists of nominees. If coupled with feature descriptions ("I prefer L to M because of property N" then such lists gain some value.
In keeping with the confidentiality rules, one of the properties of the tool is that a person will never see their own name on the list of names on which feedback is solicited. This is to avoid telling the person whether they are, or are not shortlisted. It is presumed that the individual is capable of providing feedback on themselves anyway. As noted below, this particular piece of opacity does not work. Folks can just ask around, find someone else they know who has also solicited, and usually find out if their name was on the list. (Yes, it can be argued that the confidentiality rules tell people not to relay that sort of information. But since we know from individual reports that this happens frequently, the reality seems more relevant than the theory.)
| TOC |
The process of selecting candidates for confirmation, and then working with the confirming body, is a complex and mostly confidential one. It is the nominating committee chair's job to find a process which will work, and then shepherd it through to completion.
The nominating committee used many different tools, with conversations (both voice and email) being the most important one. At various points, selection processes were applied to individuals or slates of individuals. When explicit votes were needed, preferential ballots were used, with resolution based on Condorcet voting with Instant-runoff voting (IRV) for resolving those situations where Condorcet was inconclusive. (These techniques are described in many web sites, and references were provided to the committee as part of the process of agreeing to use this resolution mechanism. I do not know how authoritative any one site is, so the reader is advised to search on their own if they are interested in more details on preferential voting resolution methods and their pluses and minuses.) This was managed through open source tools, and proved quite effective.
Candidates were selected, approved by confirming bodies, and their willingness and availability verified by a few days after the RFC specified February 22nd deadline.
| TOC |
A number of issues of various types were encountered during the nominating committee process. As a rule of thumb, it seems the best thing a nominating committee chair can do when planning the schedule is to assume things will go wrong. So push the schedule hard even when it looks like there is plenty of time. You will need it.
The following sections describe some of the issues which in my personal opinion, based on what I observed, should be kept in mind in working with the process. It is my recommendation that some of these issues be addressed by changes in the defined processes.
| TOC |
One of the first items of confusion was when it was noticed that the spirit of the exclusion from eligibility to serve on the NomCom and the letter of the law do not match. The spirit is that folks serving on bodies to which the NomCom is making appointments shall not themselves be volunteer members of the NomCom. The letter of the law lists the IAB and IESG. It does not list the IAOC.
One volunteer realized after he volunteered that as a sitting IAOC member, he should not be a volunteer member of the NomCom. So he politely asked to remove himself from the list. I removed him. But the letter of the law ought to match the intent. Until it does, NomCom chairs should keep an eye open for this potential inconsistency. If the defining documents are revised for any reason, the text there should be repaired as well.
| TOC |
As mentioned earlier, I, as chaired worked out with the confirming bodies and the liaisons the rules for the liaison participation in the NomCom. The goal of this was to enable the liaisons to provide useful input while avoiding the reported historical problems of some liaisons in some years having too much influence in the deliberation process. There is distinct ambiguity in the written rules as to the intended and desired scope of participation by the liaisons in the NomCom discussions.
As per the rules, and in order to permit the liaisons to be present for as much of the discussion as possible, the liaisons were permitted to participate in the discussions of how often we met or held phone calls, and when such activities occurred.
The terms of participation in the discussion of individuals and positions were designed to allow useful information, while preventing excess.
| TOC |
The RFCs defining the NomCom process require that all information, including the lists of nominees, who has accepted nomination, and the lists of who is under actual consideration by the nominating committee, be kept confidential. While this is an attractive rule, it has resulted in a major problem and I believe needs to be revised.
As described above, the nominating committee needs information from the community. To comply with current rules a padded set of names is sent to a portion of the community for feedback. This has multiple problems. First and foremost, it does not achieve the objective. Folks will learn if they are on the list being solicited. With the number of people who see the list, folks who want to know can find out who is on the list. (People are people, and they will talk to each other. Pretending otherwise is unrealistic. Hundreds of people do not keep a secret.) And people will determine who the pads are. Many times, no matter how hard the committee works, it will be obvious which names are padding.
As a minor point, the padding is extra work. The committee has to come up with credible names that they think won't be well known as having turned down the nomination. And then the chair really should check with those folks before using them as padding. I made some mistakes in that regard, and it caused some people difficulty.
The most important reason for removing the confidentiality of the nominees list is that the current process prevents the most important source of input. The community members who are not the leadership, the grass roots if you will, are the people who may well know important and useful information about nominees, but are not going to be solicited, much less interviewed. One of the main points of this nominating committee process is to avoid having the leadership self-appoint. Having a large portion of the community know the list of nominees, and another large portion not know the list, is counter to the goals.
My personal recommendation would be that the list of all individuals who have accepted nomination for any given position be available publicly starting 24 hours after the initial solicitation of nominees. This should include the individual, affiliation, and position, but no information as to where the nomination came from. Incumbents should be treated as anyone else in terms of listing, listing them when they have agreed to be considered as nominees. Treating the incumbents any differently from other folks would merely confuse the feedback.
| TOC |
Appointment of members to the IAB (either new members, or reappointment of existing members) is an important part of the nominating committee job. This year there was a complication in this task. As far as the nominating committee could determine, the community at large had very little understanding of what the IAB did, or what made a good or bad IAB member. There was also, for a variety of reasons, very little documentary evidence as to what individual participants in the IAB had done in that role. This was further complicated by confusion in the community, and possibly on the nominating committee, as to what actions were just a person being a good IETF member, and what actions were a person being a positive IAB member. It is presumed that the two are not interchangeable.
Part of what made this particularly complex is that the issue was not about whether the IAB is a positive part of the IETF. The community sees the IAB as an important and positive part of the IETF. But when asked to take that a step further, into the details which affect the selection process, there was significant difficulty.
| TOC |
For many very good reasons, the IETF does not have strict rules on how many people from a company may serve on the IAB or IESG (or IAOC.) At the same time, it is understood by the nominating committee that having a large number of people from one company on any given body is not a good idea.
The only reason this is even an issue is that when it comes time to decide, different members of the NomCom, and of the community, will weight this differently. That too is fine. It is merely important to understand that this is a legitimate issue for nominating committee members to consider.
There is a potential side-issue, which I would also prefer not result in an additional rule. The nominating committee rules cover the affiliations of the volunteers. They do not count the affiliation of the liaisons or the chair. It would probably be a good idea for those bodies appointing liaisons to keep an eye open for extreme cases. It would be awkward to have a committee with two volunteers, two liaisons, and a chair all from the same company. It would not violate the letter of the rules, but would certainly be seen as unfortunate.
| TOC |
One of the topics that came up this year is the policies with regard to ADs and chairs. As with other things, this is a matter where the nominating committee must judge the temper of the community. The conclusions described here are this years committees understandings of the community view. There are two loosely related issues.
The first is the question of ADs serving as working group chairs. This can occur across areas, or within an area. The sense we understood from the community is that even across areas this is not a good idea. ADs serving as chairs clearly have much more influence with their Area Director, and as such weaken the checks and balances of the system. Within their own area, even with their co-AD serving as responsible AD, this is seen as even more of a bad idea. Obviously, there are transition situations (it takes time to find new chairs) and special situations ("I thought it was done already") that complicate any hard rule. But as a sense of the community, this seemed to this committee to be a bad idea. It is not clear where, other than this sort of document, these observations about "bad practices" can be recorded to assist future activities.
A much subtler and more nuanced situation is the way that various ADs go about picking chairs for working groups. It is very common for ADs to want to pick chairs that they can work with. It makes life MUCH easier. This does however have a tendency to center the pool of chairs in an area around a few small groups. Some ADs have used public calls for nominations as a way to counter this trend. Such public calls often bring forward names that the AD would not have considered who are strong, capable individuals. Sometimes even more complex canvassing for good chairs is needed, particularly when the working group has a complex set of constraints it needs to work within.
| TOC |
There are no IANA Considerations in this document, as no computer protocols are discussed, and no code points assigned. This section may be removed by the RFC Editor.
| TOC |
This is a report on an activity that occurred. Every effort has been made to respect the mandatory confidentiality aspects of that activity. Some of the proposals in this document would change those properties, and such effect are clearly an important part of the consideration of such changes. It is the author's view that the trade-off would be to the benefit of the community.
The NomCom this year followed the practices of the last several years with regard to managing information confidentiality. The archives of the comments, questionnaires, and other email to the committee are stored encrypted on a server manage by Henrik, with the decryption key known to the committee members. Distribution of that private key used a private password protected web site, with the password provided to committee members by phone. If the committee can be organized before the second face-to-face IETF meeting of the year, the decryption key can be distributed manually at that meeting.
| TOC |
| TOC |
| [RFC3777] | Galvin, J., “IAB and IESG Selection, Confirmation, and Recall Process: Operation of the Nominating and Recall Committees,” BCP 10, RFC 3777, June 2004 (TXT). |
| TOC |
| [RFC3797] | Eastlake, D., “Publicly Verifiable Nominations Committee (NomCom) Random Selection,” RFC 3797, June 2004 (TXT). |
| TOC |
| Joel M. Halpern (editor) | |
| Ericsson | |
| P. O. Box 6049 | |
| Leesburg, VA 20178 | |
| US | |
| Email: | jhalpern@redback.com |