On Jul 30, 2013, at 6:10 PM, Michael Richardson <mcr+ietf@xxxxxxxxxxxx> wrote: > > Keith Moore <moore@xxxxxxxxxxxxxxxxxxxx> wrote: > >> Rooms are set up not to facilitate discussion, but to discourage it. The >> lights are dim, the chairs are facing forward rather than other participants, >> the projector screen (not the person facilitating a discussion, even if someone >> is trying to facilitate a discussion) is the center of attention. The chairs >> are set so close together and with so few aisles that it's hard for most of the >> attendees to get to the mics. The "microphone discipline" which was intended >> to facilitate remote participation ends up making discussion more difficult for >> everybody who has paid to be on site. > > I think that these physical things are something that we can do some > experiments about. > >> Well, please excuse my candor, but f*ck habit. We can't be effective >> engineers if we let bad habits continue to dictate how we work. > > I agree. > >> For 80% of most WG meetings, the lights should be bright, the participants >> should face each other. If there's a person facilitating the discussion that >> person should be the center of attention. If we're going to use microphones, >> the rooms should be set up to allow everyone in the room to have easy access to >> them. We should have several microphones, again facing each other, so that >> several people can have a conversation without everyone having to queue up. > > Can we please try this in Vancouver? > This would work especially well for BOFs. > Maybe we can start there. > Chairs will need training as *facilitators* The IETF is not tasked with inventing "discussion" or even "collaborative design". This things already exist, and we can maybe look around and see what works in other places. Many of us work for vendors or other groups where designs are created in collaboration. At least where I work now, we don't do design by facing forward while someone presents. The usual way is 2-4 people standing next to a whiteboard drawing stuff. Photographing the whiteboard with a smartphone has become a new addition, and using a flip-chart is a fine variation. There are the larger design meetings (usually attended by a manager, some QA people, a technical documenter, and other "stakeholders"). Those tend to have some slides, but the whiteboard is still there and people generally sit around a table. Even those larger meetings rarely go over 10-12 people. This experience doesn't transfer well to working groups with 50-100 participants, some of which are remote. So what is one to do? Well, one option is to move most of the discussion out of these "main events". Sure, you can still have small design sessions at an IETF meeting, but I think the bulk of the documents that we work on have less than 12 people actively working on them (yes, http/2.0 is an outlier here). These people can meet often virtually, and there are several tools that allow these virtual interim meetings that even allow you to shared a whiteboard. The results of such meetings can and should be reported to the WG mailing list for feed-back, and then we can have real presentations at the face-to-face meetings, where we solicit comments by those who are not part of the core group, and are not following closely. I just don't believe in effective discussion among a group of 50-100 people. The best we can hope for is a Q&A session with the design team. If you want another example of an ineffective discussion in a group of 100+ people, just take a look at your local parliament / congress. There's a reason why all of them have formed committees, and why the executive branch has usurped much of their power in many countries. Trying to force such a body to effective discussion will not work. Yoav