RFC 
 3683 
 TOC 
Network Working GroupM. Rose
Request for Comments: 3683Dover Beach Consulting, Inc.
BCP: 83February 2004
Category: Best Current Practice 


A Practice for Revoking Posting Rights to IETF mailing lists

Status of this Memo

This document specifies an Internet Best Current Practices for the Internet Community, and requests discussion and suggestions for improvements. Distribution of this memo is unlimited.

Copyright Notice

Copyright © The Internet Society (2004). All Rights Reserved.

Abstract

All self-governing bodies have ways of managing the scope of participant interaction. The IETF uses a consensus-driven process for developing computer-communications standards in an open fashion. An important part of this consensus-driven process is the pervasive use of mailing lists for discussion. Notably, in a small number of cases, a participant has engaged in a "denial-of-service" attack to disrupt the consensus-driven process. Regrettably, as these bad faith attacks become more common, the IETF needs to establish a practice that reduces or eliminates these attacks. This memo recommends such a practice for use by the IETF.


 RFC 
 3683 
 TOC 

Table of Contents

1.  Introduction
2.  A Revocation Practice
3.  Acknowledgements
4.  Security Considerations
5.  Normative References
Appendix A.  Q & A
§  Author's Address
§  Intellectual Property and Copyright Statements




 TOC 

1.  Introduction

All self-governing bodies have ways of managing the scope of participant interaction. For example, deliberative assemblies often employ "rules of order" for determining who gets to speak, when, and for how long. Similarly, there is widespread agreement in so-called "liberal" societies that the right to free speech is not absolute, e.g., political speech is given more leeway than commercial speech, and some forms of speech (e.g., egregious libel or incitement to violence) are considered unacceptable.

The IETF uses a consensus-driven process for developing computer-communications standards in an open fashion. An important part of this consensus-driven process is the pervasive use of mailing lists for discussion. Unlike many other organizations, anyone may post messages on those IETF mailing lists, and in doing so, participate in the IETF process. Historically, this approach has worked very well in the IETF, as it fosters participation from a wide range of stakeholders. (For the purposes of this memo, the term "IETF mailing list" refers to any mailing list functioning under IETF auspices, such as the IETF general discussion list, or a working group or design team mailing list.)

Notably, in a small number of cases, a participant has engaged in what amounts to a "denial-of-service" attack to disrupt the consensus-driven process. Typically, these attacks are made by repeatedly posting messages that are off-topic, inflammatory, or otherwise counter-productive. In contrast, good faith disagreement is a healthy part of the consensus-driven process.

For example, if a working group is unable to reach consensus, this is an acceptable, albeit unfortunate, outcome; however, if that working group fails to achieve consensus because it is being continuously disrupted, then the disruption constitutes an abuse of the consensus-driven process. Interactions of this type are fundamentally different from "the lone voice of dissent" in which a participant expresses a view that is discussed but does not achieve consensus. In other words, individual bad faith should not trump community goodwill.

Guidelines have been developed for dealing with abusive behavior (c.f., Section 3.2 of [RFC2418] (Bradner, S., “IETF Working Group Guidelines and Procedures,” September 1998.) and [RFC3005] (Harris, S., “IETF Discussion List Charter,” November 2000.)). Although not exhaustive, examples of abusive or otherwise inappropriate postings to IETF mailing lists include:

In practice, the application of those guidelines has included the temporary suspension of posting rights to a specific mailing list. If necessary, the length of the suspension has been increased with each successive suspension. In many cases, applying those guidelines will produce the desired modification in behaviour. However, when those guidelines fail to provide the desired modification in behaviour, more drastic measures should be available to reduce or eliminate these attacks' impact on the IETF process.

This document describes one such drastic measure.



 TOC 

2.  A Revocation Practice

Please refer to [RFC2119] (Bradner, S., “Key words for use in RFCs to Indicate Requirement Levels,” March 1997.) for the meaning conveyed by the uppercase words in this section.

As a part of its activities, the Internet Engineering Steering Group (IESG) makes decisions about "actions". Typically, an action refers to the publication of a document on the standards-track, the chartering of a working group, and so on. This memo recommends that the IESG also undertake a new type of action, termed a PR-action ("posting rights" action).

A PR-action identifies one or more individuals, citing messages posted by those individuals to an IETF mailing list, that appear to be abusive of the consensus-driven process. If approved by the IESG, then:

Once taken, this action remains in force until explicitly nullified and SHOULD remain in force for at least one year.

One year after the PR-action is approved, a new PR-action MAY be introduced which restores the posting rights for that individual. The IESG SHOULD consider the frequency of nullifying requests when evaluating a new PR-action. If the posting rights are restored the individual is responsible for contacting the owners of the mailing lists to have them restored.

Regardless of whether the PR-action revokes or restores posting rights, the IESG follows the same algorithm as with its other actions:

  1. it is introduced by an IESG Area Director (AD), who, prior to doing so, may choose to inform the interested parties;
  2. it is published as an IESG last call on the IETF general discussion list;
  3. it is discussed by the community;
  4. it is discussed by the IESG; and, finally,
  5. using the usual consensus-based process, it is decided upon by the IESG.

Of course, as with all IESG actions, the appeals process outlined in [RFC2026] (Bradner, S., “The Internet Standards Process -- Revision 3,” October 1996.) may be invoked to contest a PR-action approved by the IESG.

Working groups SHOULD ensure that their associated mailing list is manageable. For example, some may try to circumvent the revocation of their posting rights by changing email addresses; accordingly it should be possible to restrict the new email address.

Finally, note that the scope of a PR-action deals solely with posting rights. Consistent with the final paragraph of Section 3.2 of [RFC2418] (Bradner, S., “IETF Working Group Guidelines and Procedures,” September 1998.), no action may be taken to prevent individuals from receiving messages sent to a mailing list.



 TOC 

3.  Acknowledgements

The author gratefully acknowledges the contributions of: Brian Carpenter, Jim Galvin, Jeff Haas, Ted Hardie, Russ Housley, Thomas Narten, Jon Peterson, Margaret Wasserman, and Bert Wijnen.



 TOC 

4.  Security Considerations

This memo deals with matters of process, not protocol.

A reasonable person might note that this memo describes a mechanism to throttle active denial-of-service attacks against the consensus-driven process used by the IETF.



 TOC 

5. Normative References

[RFC2418] Bradner, S., “IETF Working Group Guidelines and Procedures,” BCP 25, RFC 2418, September 1998 (HTML, XML).
[RFC3005] Harris, S., “IETF Discussion List Charter,” BCP 45, RFC 3005, November 2000 (TXT).
[RFC2119] Bradner, S., “Key words for use in RFCs to Indicate Requirement Levels,” BCP 14, RFC 2119, March 1997 (HTML, XML).
[RFC2026] Bradner, S., “The Internet Standards Process -- Revision 3,” BCP 9, RFC 2026, October 1996.


 TOC 

Appendix A.  Q & A

Q:
Isn't a year too long?
A:
No.
An initial PR-action is not undertaken lightly. It is approved only after a period of substantive consideration and community review. If a PR-action is approved, then this indicates that a serious situation has arisen.
Q:
Why not require one PR-action per IETF mailing list?
A:
To do so would enable a prolonged series of denial-of-service attacks.
If someone is poorly-behaved on one IETF mailing list, but well-behaved on another, then the maintainer for the second IETF mailing list needn't revoke posting rights. However, the more likely scenario is that someone who behaves poorly on one IETF mailing list is unwilling to be well-behaved on any IETF mailing list.
Q:
Should the initiation of a PR-action come from outside the IESG?
A:
Informally, sure; formally, no.
Under the IETF's consensus-driven process, IESG actions are always formally initiated by an IESG Area Director (AD). In practice, the motivation for an IESG member to initiate an action almost always comes from outside the IESG. For example, when a working group (WG) reaches consensus on a document, the WG chair informs the relevant AD that the document is ready for the AD to consider it for a document action. In the case of this document — an IETF individual submission — the author will iteratively circulate the document for wide discussion and make revisions. At some point, the author will contact an AD and ask for a document action to publish this document as a Best Current Practice (BCP).
Q:
Is this censorship?
A:
Only if you believe in anarchy.
What is important is that the rules surrounding PR-actions exhibit the same properties used by the rest of the consensus-based process.
Q:
C'mon! You really are a closet fascist.
A:
No, I'm a libertarian.
Frankly, I would prefer that people behave reasonably and act in good faith. Since my first involvement with the IETF (nee GADS, circa 1983), everyone understood that reasonable behavior was a good thing. After 20 years, I regret to inform you that this step is inevitable.


 TOC 

Author's Address

  Marshall T. Rose
  Dover Beach Consulting, Inc.
  POB 255268
  Sacramento, CA 95865-5268
  US
Phone:  +1 916 483 8878
Email:  mrose@dbc.mtview.ca.us


 TOC 

Full Copyright Statement

Intellectual Property

Acknowledgment