ietf
[Top] [All Lists]

Re: Interim meetings - changing the way we work

2015-02-25 04:40:46
Hi Tom,

In the WGs for which interim meetings are hold, there is really an increased speed of development. I can see this in NETMOD for example. So I've been encouraging more interim meetings, when we have the right audience in the call.

While we transition to more interim meetings, we have to admit that different WGs uses different tools (webex, github, svn, etherpad, irc). I don't believe that imposing the interim meeting tools is right. We should let each community chose. Coming back to NETMOD (which is missing some entries at http://www.ietf.org/meeting/interim/proceedings.html), it works with SVN. Everything is documented at https://svn.tools.ietf.org/svn/wg/netmod/yang-1.1/, as communicated on the mailing list. However, you are right, we should follow the process, and go the extra mile to complete the information at http://www.ietf.org/meeting/interim/proceedings.html

Final thought. I believe that some of our tooling will need to adapt anyway for interim meeting management. For example, I missed some interim meetings in my WGs based on the simple rule: if it's not in my calendar, it doesn't exist. The https://www.ietf.org/meeting/interim.html format is not very helpful for that. We need a way to easily schedule, evaluate conflicts, and download the .ics ... in a type of calendar view.

Regards, Benoit
There has been a marked increase in the number of  interim meetings.
Using
http://www.ietf.org/meeting/interim/proceedings.html
as a guide, there were
18 in 2011
35 in 2012
45 in 2013
84 in 2014
13 in January 2015 alone.

With them comes a change in the way of working, perhaps rendering some
of our practices historic.

Of the 84 meetings listed for 2014, 21 left no other trace on the IETF
web site, no Agenda, no Minutes, no Proceedings.  Perhaps the WG
provided no materials, perhaps they did not happen; sometimes a
cancellation notice is apparent in the WG List Archives, other times
not.

Of the 63 that have left a trace, 6 produced no Minutes but did produce
slides or recordings and so presumably happened.

Of the 57 that produced Minutes, 18 produced no Agenda while in 13
cases, the Minutes contained no list of Attendees (goodbye Blue
Sheets?).

Only 26 meetings left a complete record, of Agenda, Minutes and
Attendees.

The meetings encompassed 30 Working Groups, of which 16 met once, 14
more than once, with one WG meeting 8 times.

What is more subjective is that, with Virtual Interims, increasingly the
only kind, there is a tendency for the WG Mailing List to no longer
provide a record of discussions, choices, consensus.  For example, they
may make greater use of github so that the minutes record a discussion
of options 1, 2 and 3 for Issue 29 with no indication of what the issue
or options are; a while later, they may record an update to option 3 so
it would now seem impossible to know what was discussed at the earlier
meeting.

Even with the better minutes, they never give the same sense as posts to
a mailing list of who was or was not in favour and how strong their view
was.

Of course, we still have WG Last Calls on the list but if at a future
date, an AD or GenArt reviewer wants to look back and see what options
were discussed and  how rough the consensus was, well, it may be
impossible.

A post in another thread recently said

I do think that the increased significance of meetings
in IETF participation (and here, I'm not talking about
things like nomcom but about significance to our technical
work) is a problem, both because it tends to marginalize
people who can't come to meetings and because it slows
work down.
Well, I disagree about slowing the work down but certainly agree with
the marginalisation, that WGs holding multiple Interims may tend to
develop an in-crowd of those that can participate with the world at
large only seeing the end result without knowing how it was arrived at
by whom.

Tom Petch

.