Internet Engineering Task Force (IETF)
Request for Comments: 6771
Category: Informational
ISSN: 2070-1721
L. Eggert
NetApp
G. Camarillo
Ericsson
October 2012

Considerations for Having a Successful "Bar BOF" Side Meeting

Abstract

New work is typically brought to the IETF by a group of interested individuals. IETF meetings are a convenient place for such groups to hold informal get-togethers to discuss and develop their ideas. Such side meetings, which are not reflected in the IETF meeting agenda and have no official status, are often half-jokingly referred to as "bar BOF" sessions to acknowledge that some of them may eventually lead to a proposal for an official IETF BOF ("birds of a feather" session) on a given topic.

During recent IETF meetings, many such "bar BOF" get-togethers have been organized and moderated in ways that made them increasingly indistinguishable from official IETF BOFs or sometimes even IETF working group meetings.

This document argues that this recent trend is not helpful in reaching the ultimate goal of many of these get-togethers, i.e., to efficiently discuss and develop ideas for new IETF work. It encourages the organizers to consider the benefits of holding them in much less formal settings and to also consider alternative means to develop their ideas. This document also recommends that the community abandon the term "bar BOF" and instead use other terms such as "side meeting", in order to stress the unofficial nature of these get-togethers.

Status of This Memo

This document is not an Internet Standards Track specification; it is published for informational purposes.

This document is a product of the Internet Engineering Task Force (IETF). It represents the consensus of the IETF community. It has received public review and has been approved for publication by the Internet Engineering Steering Group (IESG). Not all documents approved by the IESG are a candidate for any level of Internet Standard; see Section 2 of RFC 5741.

Information about the current status of this document, any errata, and how to provide feedback on it may be obtained at http://www.rfc-editor.org/info/rfc6771.

Copyright Notice

Copyright © 2012 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 (http://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.

1. Introduction

A typical IETF meeting is full of sessions of different kinds. In addition to official IETF and IRTF sessions listed in the meeting agenda (such as working and research group meetings, area meetings, or plenaries), many other unofficial meetings take place. These include meetings between IETF participants from one organization or company, design team meetings, Internet-Draft editing sessions, interoperability testing, directorate lunches, and many others.

Some of these unofficial get-togethers are organized by individual participants with a common interest in initiating new IETF work of some kind. New IETF work often fits into an existing working group and does not require an official "birds of a feather" (BOF) session [RFC5434] to determine community consensus. Nevertheless, the phrase "bar BOF" has commonly been used in the community when talking about such informal get-togethers that are held to discuss potential new work. [RFC4677] (which has been obsoleted by [RFC6722]) characterizes a "bar BOF" as

an unofficial get-together, usually in the late evening, during which a lot of work gets done over drinks. Bar BOFs spring up in many different places around an IETF meeting, such as restaurants, coffee shops, and (if we are so lucky) pools.

During recent IETF meetings, "bar BOFs" have become increasingly indistinguishable from official IETF BOFs or sometimes even IETF working group meetings. The symptoms of this trend are unofficial "bar BOFs" that are held in regular IETF meeting rooms with classroom-style seating, agendas with lengthy slide presentations, use of microphone lines, and even formal consensus calls. And, perhaps most importantly, such meetings have a distinct lack of drinks.

This document argues that this trend is not helpful in reaching the ultimate goal of many of these get-togethers, i.e., to brainstorm about a technical topic that may eventually lead to new IETF work. It encourages the organizers of these unofficial get-togethers to consider the benefits of holding them in much less formal settings.

This document also recommends that the community abandon the term "bar BOF". The distinction between a BOF, i.e., an official IETF activity, and a "bar BOF", i.e., an unofficial get-together, is lost on many IETF participants, especially newcomers. The similarity in terms has even caused confusion to the point where some participants believe that a "bar BOF" is a required step in the IETF process in order to apply for an official BOF, which is obviously false. For these reasons, the remainder of this document will use the term "side meeting" instead and recommends that the community do the same, in order to stress the unofficial nature of these get-togethers.

Before going into more detailed advice on how to hold side meetings, it is important to remember that many participants are extremely busy during an IETF meeting. Although having a side meeting to discuss an idea in an informal face-to-face setting is attractive, the scheduling of such meetings is very difficult and needs to happen weeks, if not months, prior to the meeting itself. Conference calls, email discussions, wikis, jabber group chats, and other ways for interacting are also effective at developing ideas and easier to schedule.

2. How to Invite

A good rule of thumb is that a side meeting to discuss and develop a proposal for new IETF work should include the necessary participants to achieve that purpose and no more. Smaller meetings are usually more successful than larger meetings.

Hence, it is often useful to limit attendance carefully. Publicly broadcasting an announcement for a side meeting on a particular topic, e.g., on an IETF mailing list, is therefore not usually a good method of inviting the desired set of participants.

One reason is that if the announcement happens to attract a large response, the logistics of organizing a side meeting for a larger group quickly becomes very difficult. Small groups fit comfortably around a table at a bar or a restaurant or can find a quiet corner in an IETF hallway for a discussion. Larger groups require dedicated meeting facilities, which are limited during IETF meetings, and they generally require much more careful planning in order to get work done.

When publicly announcing a side meeting, it is often not even possible for the organizers to determine how large the resulting get- together will be, forcing them to over-provision for the "best case" of a substantial attendance, even in cases where this turns out to be not necessary. And even when a large group comes together, it often mostly consists of "tourists". Tourists usually do not actively participate in the get-together at all, or they participate with an intent to learn about a topic, which can derail a planned discussion of specific issues and turn it into a tutorial. The attendance of tourists requires finding a larger room and makes the interactions between the active participants more cumbersome, e.g., because microphones need to be used in larger rooms. There are times to expose new ideas to a broader community, but think carefully before publicly announcing a side meeting.

So while publicly announcing a side meeting can be useful in order to gather interested people for a discussion, it often makes sense to still limit attendance. For example, an announcement could say, "We have a table reserved at restaurant X for Y people. If you are interested in attending, please briefly explain how you will contribute to the discussion we are planning to have". If more than Y people respond, the organizers make a selection.

Selecting or specifically inviting IESG or IAB members is not necessary and may not even be advisable in many cases. Some ideas need time to form before they result in anything cohesive, and a side meeting is a good time to develop new ideas. It is usually most useful to approach Area Directors (ADs) and IAB members for comments after an idea has solidified enough so that an elevator pitch can be given. Also, it should be clear that if an AD or IAB member attends a side meeting, it is not necessarily a show of support. They may simply be interested or often may be concerned or troubled with some aspect of the potential work and relation to existing work. On the other hand, when an AD or IAB member declines to attend a side meeting, that is usually not a sign of disinterest or disapproval -- these people have busy schedules, especially during an IETF week.

In the initial stages of developing a proposal for new IETF work, the ability for interested and experienced participants to brainstorm is tremendously important. Brainstorming is facilitated by direct, interactive, and high-bandwidth discussions. This is clearly much more easily achieved in a smaller setting, where half-baked ideas can be dissected and developed. This is often not possible in a larger group. Even worse, a badly run large meeting can sometimes "poison the waters" for a proposed idea by convincing the broader community that the proposal is confused, not ready, or otherwise uninteresting.

Another reason to discuss new work proposals in smaller groups is scope creep, i.e., the tendency of an initially rather tightly scoped area of new work to expand, because people will argue that whatever the initial topic was, it should be expanded to include their particular item of interest. This is harder to control in larger groups. Keeping the scope of new work items narrow is important, because eventual chartering decisions are often much more difficult for larger items of new work than for smaller ones.

It is important to understand that in the IETF, proposals for new work are judged based on their technical merits and on whether there is enough energy and interest in the community to complete the work in a timely manner. This happens in the relevant working group, if one exists, or else during an official BOF session. How many warm bodies fill a room during an unofficial side meeting has no influence on this decision and is not a good metric for reporting interest in a topic to the community or to employers. Discussions about new work are often controversial, and people will show up just to watch the fireworks, learn about a new topic, or make sure the new work does not interfere with work they are already pursuing, without being interested contributing in some way to the actual proposal itself.

Some side meetings are organized to discuss a topic that is also being discussed in an existing working group, either before or after the working group itself meets. Some working groups call these side meetings "ad hoc sessions". The fact that a side meeting is organized by a chair or key participant of a working group in order to discuss topics related to the working group does not make it any more official than other side meetings. An "ad hoc session" is not an official working group session, and no decisions relevant to a working group can be made. Working group consensus can only be established during official sessions or on the mailing list [RFC2418].

3. Where to Meet

As the colloquial name "bar BOF" implied, such side meetings are traditionally held in bars or restaurants. Recently, there has been a distinct shift towards holding such get-togethers in regular IETF meeting rooms. One reason for this trend has been discussed in Section 2, namely, that an uncontrolled broadcast announcement requires over-provisioning of facilities.

A second reason for this trend is that some participants, e.g., non- native English speakers or participants with hearing difficulties, find it difficult to interact or follow a discussion in noisy environments, such as restaurants and especially bars. The organizers of side meetings are encouraged to take this factor into consideration when finding a meeting place. Quiet restaurants are not hard to find, and many offer private dining rooms at no extra charge for larger parties.

   A likely third reason why side meetings are increasingly held in IETF
   rooms is that the booking of such a room currently requires approval
   by an Area Director.  The reason for this practice is simply to make
   sure that IETF-paid rooms are used for meetings that are in the
   widest sense IETF-related.  However, the approval of a room request
   for a side meeting has been known to sometimes be reported as Area
   Director "support" for the topic of the meeting to the community or
   to employers.  No such support is expressed or implied when Area
   Directors approve room requests!  Many routinely say "yes" to every
   incoming request as long as there are meeting rooms available (and
   there are typically lots of meeting rooms available outside of normal
   working group meeting slots).

Holding side meetings in IETF meeting rooms does not make them any more official or valid than get-togethers that happen in other places. Participants have recently begun to list the times and locations of some side meetings on a wiki page, but that does not make them part of the official IETF agenda or otherwise change their unofficial status.

IETF meeting rooms clearly do not provide the most supportive environment for side meetings that require brainstorming on a new technical proposal. One reason is that the classroom-style seating often present in IETF meeting rooms tends to spread people out in rows, all facing towards a front presenter, which is good for presentations but bad for discussion. Because IETF meeting rooms tend to be large and people have a natural tendency to spread out, holding a meeting in one often requires microphone use, which is cumbersome, slows a discussion down, and leads to "question-answer" dialogs between two people, which is much less effective than a group discussion around a restaurant table.

Another reason is more pragmatic. Because the organizers of unofficial get-togethers can only use IETF meeting rooms during times when they are not otherwise in use, such side meetings often happen during breakfast, lunch, dinner, or later in the evening. This prolongs the time during which IETF participants are stuck in the same rooms they're stuck in for the rest of the day, and it prevents them from having a regular and at least somewhat relaxed meal.

Anecdotal evidence exists that at least one Area Director has not been able to set foot outside the IETF hotel for a stretch of several days during IETF 77. (IETF 77 was held in Anaheim, CA, and the food options in and near the hotel were, let's say, of severely limited quality.) It is unlikely that participants in the consequential mental and bodily state will make productive contributions to a side meeting or, in the case of Area Directors, will be extremely receptive towards new work proposals.

Food, drink, and a relaxed atmosphere in which to have a discussion are an essential part of a successful side meeting, because they often need to happen during meal times. IETF meeting rooms offer neither.

4. How to Meet

Several of the recent side meetings that were held in IETF meeting rooms emulated official IETF meetings to a degree that made them indistinguishable from a regular working group meeting for the average IETF attendee. This included detailed agendas, lengthy presentations, organizers who refer to themselves as "bar BOF chairs", emulating blue sheets (see Section 4.5 of [RFC4677]), and even hums and other consensus calls (see Section 5.2 of [RFC4677]).

It is not clear as to why this has been happening. One attempt at an explanation may be that holding a get-together in an IETF room and having the organizers behave like chairs behave during regular IETF sessions is causing a Pavlovian stimulus in the attendees. Another explanation attempt is that an IETF meeting room simply does not allow many other forms of discussion. Finally, some organizers may find the process to apply for an official BOF too complex or troublesome (and probably rightfully so) and so decide to simply mimic one, or they had applied for an official BOF, got turned down, and then decided to hold the same meeting as a side meeting.

Whatever the reason for this development, it is reasonably obvious that running a side meeting with a focus on making quick progress on a technical proposal in a way that emulates running a working group session is not very productive. Working group sessions follow certain procedures due to larger audiences, the need to establish formal consensus, etc., that a side meeting can do without.

Having side meetings mimic working group meetings is also confusing to attendees. In at least one case, some side meeting participants believed that they were attending an actual working group meeting, and incorrect press announcements were generated. When side meetings take place at restaurants or elsewhere away from IETF meeting rooms, the chance for confusion is much lower.

Because the reasons for organizing such a get-together are diverse, this section is not making more specific suggestions, other than to note that meeting outside of an IETF meeting room is likely going to shift the dynamics sufficiently so that better interactions and results become possible.

5. When to Meet

Side meetings are often scheduled following IETF evening plenaries, which sometimes end before the time indicated on the meeting agenda but have in the past also ended much later. It is therefore useful to avoid scheduling side meetings that follow IETF plenaries at a fixed time. Instead, it is recommended to schedule them relative to the end of the plenary, i.e., "X minutes after the end of the plenary". That way, attendees do not need to wait around if a plenary finishes early and do not need to leave a plenary should it run late.

Section 3 of [RFC5434] raises the issue that it is essential to understand all angles of a given problem for which IETF work is proposed. This means that input from the community that can be found at IETF meetings is not all that should be considered. It can be argued that input from other communities -- operator, research, regulatory, etc. -- is at least as important. Hence, organizers should consider the value of holding side meetings at venues where such input can be more easily gathered, such as operator fora (RIPE, NANOG, etc.), research conferences, or other events.

6. Conclusions

Side meeting organizers are encouraged to rekindle the original spirit behind them and organize them outside IETF meeting rooms, at venues with food and drink, for smaller groups, and in a way that does not needlessly mimic the way official IETF sessions are conducted.

It can often be useful to discuss proposals for new IETF work face- to-face in an informal setting, but conference calls, email discussions, wikis, and other means for interactions are also effective at developing ideas, especially given the scheduling difficulties when busy individuals are involved during an IETF meeting.

Finally, it is important to remember that all side meetings during an IETF week are purely informal and have no official status whatsoever.

7. Security Considerations

A security AD pointed out that people have been known to forget their laptops after side meetings held in real bars. The organizers of side meetings should therefore remind any attending security ADs (and possibly others) to take their belongings with them after the side meeting ends or the bar closes, whichever happens first.

8. Acknowledgments

The name and title of this document have been chosen to resemble those used by Thomas Narten for his guidelines document on holding a successful BOF [RFC5434], as a sign of appreciation for a document that has proven to be invaluable many times over.

Several folks provided feedback and input on this document, including Jari Arkko, Fred Baker, Scott Bradner, Ben Campbell, Jorge Contreras, Spencer Dawkins, Ralph Droms, Wesley Eddy, Frank Ellermann, Adrian Farrel, Stephen Farrell, David Harrington, Russ Housley, Cullen Jennings, John Klensin, Al Morton, Robert Sparks, and Dan Wing.

Lars Eggert was partly funded by [TRILOGY], a research project supported by the European Commission under its Seventh Framework Program.

9. Informative References

   [RFC2418]  Bradner, S., "IETF Working Group Guidelines and
   
              Procedures", BCP 25, RFC 2418, September 1998.
   
   [RFC4677]  Hoffman, P. and S. Harris, "The Tao of IETF - A Novice's
              Guide to the Internet Engineering Task Force", RFC 4677,
              September 2006.
   
   [RFC5434]  Narten, T., "Considerations for Having a Successful Birds-
              of-a-Feather (BOF) Session", RFC 5434, February 2009.
   
   [RFC6722]  Hoffman, P., "Publishing the "Tao of the IETF" as a Web
              Page", RFC 6722, August 2012.
   
   [TRILOGY]  "Trilogy Project", <http://www.trilogy-project.org/>.

Authors' Addresses

   Lars Eggert
   NetApp
   Sonnenallee 1
   Kirchheim  85551
   Germany
   
   Phone: +49 151 12055791
   EMail: lars@netapp.com
   URI:   http://eggert.org/
   
   Gonzalo Camarillo
   Ericsson
   Hirsalantie 11
   Jorvas  02420
   Finland
   
   EMail: Gonzalo.Camarillo@ericsson.com