Mailing Lists: What works, what doesn’t?

(Kathleen’s opinions, not an IESG statement of any sort.)

The IETF has had another lively discussion about mailing list usage in the ietf@ietf.org mailing list, followed by a long plenary debate on how to make it more useful to the community.  Thanks to those of you who stuck out the long discussion Wednesday night, providing insight and ideas to improve from the current state.  These discussions can be painful, but necessary to figure out how we evolve with time. 

A few things seemed apparent during the session:

  • IETF participants in the ietf@ietf.org mailing list are passionate and appreciate use of that forum.   The number of active users is relatively small for the size of our community.
  • On the other hand, many are turned away from the list, seeing it as a time suck, high noise to signal ratio, and too aggressive.  These folks are very unlikely to join again.
  • The list is not an ideal venue for newcomers.  Moderation might help, but changes in culture of the list would take a year or more to be effective and noticeable to others.
  • Younger newcomers see mailing lists as archaic; perhaps a focus on change to the list to attract newcomers won’t meet our goals?
  • Ideally, feedback during the last call process would be increased because of discussions on the ietf@ietf.org mailing list.  For the most part, that is not happening.  Perhaps that is okay for at least some drafts, demonstrating that drafts are good at that stage of the process.

What are we trying to accomplish?

  • Provide a forum for IETF-wide discussions on relevant topics including last call comments, proposed policy and process changes.
  • Ensure discussion forums exist and are archived, supporting transparency goals of the IETF.
  • Increase exposure for soon-to-be published drafts for both reviews and awareness.
  • Improve the quality of Internet standards, which requires a culture that fosters retention of newcomers.

The IETF discussion list is not meeting these goals.  Many find the IETF list painful, so they avoid it.  We still need the list as some find it very helpful and enjoy working in mailing lists.  Many in the IETF community contributed helpful comments during the Toronto Administrative Plenary, and several possibilities emerged.  What if we continued use of the mailing list for community-wide discussions to avoid any controversy and experimented with new communication methods to serve the other purposes, that currently do not work well through the list?  We want to attract and retain newcomers and at least the younger ones think the use of mailing lists is archaic.  Complementary to discussion forums, we could consider the use of new communication methods may also increase awareness of IETF protocols and thus the understanding of relevance to the larger community (twitter, see below).  Many of us, including me, are quite comfortable with the datatracker and existing tools, the following suggestions are not necessarily targeted towards those comfortable with the current set of tools.

Possible experiments:

  • Create a twitter account to announce both working group and IETF-wide last calls.  Those interested can either look at the twitter feed once a week to make sure they didn’t miss something, while those that follow it may choose to retweet drafts relevant to their followers.  Discussions would be directed to mailing lists or another tool as appropriate, and twitter use would be strictly a marketing tool.
  • Create an organized site to maintain a current list of drafts in working group and IETF last call.  The site should be modern and include social aspects as to attract newcomers and experts in specific areas who may not necessarily be regular IETF attendees (yet).  Ideally, the site would include easy ways to contribute comments, even to support comments posted by others or provide a simple statement of support for the work.  There should be one place to look at all comments during last call for the draft editors, shepherds, and others.  If we experimented with such a model, to Adrian Farrel’s point, we would need to ensure ease of commenting and debating comments in a structured and traceable way.

Authenticated edit access with traceability for changes (add, remove, etc.) on the new services would be necessary, to prevent the tools from becoming as ‘entertaining’ (and ultimately ineffective) as the use of Etherpad in the Toronto Administrative plenary. 😉

Kathleen Moriarty, Security Area Director, IETF