Hi. Since this has, by last count, gone on for four days and 45 messages, some quick comments from someone who was in Montreal (and didn't touch Jabber or Meetecho) but who is usually remote these days (in no particular order): (1) Using the Meetecho remote queue would be a better substitute for the Jabber "scribe" if those who are leading meetings consistently followed the associated notification, paid attention to them, and treated such requests fairly wrt in-room participants. In practice, while I think it is gradually getting better, it often doesn't happen: many times that queue is either ignored completely or gotten to only when all in-room queues have drained. Sometimes the latter is after the "mic line" is closed, which means that the remote comments or questions are either locked out entirely or simply aren't injected into the discussion until after they cease to be timely. This is a meeting chair training, management, and focus problem, not something that can be done with technology, unless we want to modify the interfaces to be _really_ intrusive. A Jabber scribe coming to the microphone is often still not as timely as it should be for remote people to participate actively in a controversial discussion, but it works much better than a Meetecho remote queue that is being ignored. And, fwiw, the problem is not just a rogue or careless WG chair or two -- in the last few years, one of the "pay no attention to remote queue requests" offenders has often been the plenaries. (2) As someone else mentioned, following a live session and a chat room simultaneously can be rather difficult. That is as much a matter of the attention and skills of the remote participant than it is a matter of technology. Meetecho's (IMO, very clever and useful) mode-switching arrangements may pose difficult choices for the remote participant in terms of what to look at. It may be impossible to optimize simultaneously across tradeoffs: some people with excess bandwidth, some with a shortage; some in private spaces and some in busy rooms; some on small laptops, some on desktops with large screens (plural). One thing that makes it possible to follow Jabber under those circumstances is that it is serial -- the most recent activity is always at the bottom/end of the transaction list. (Alternate IM applications would be no better but would probably be harder to get, and say, used to before they keep being "improved".) Etherpad (and equivalent) have many advantages and useful features, but asking someone who is already challenged by keeping up with the Jabber feed and the meeting in real time to, instead, keep track on a pad that is being regularly modified in different places... well, others may be much better at that sort of thing than I am but I'd think it would be a major setback. (3) Another of the issues with accommodating different types of remote setups is that, even if bandwidth is adequate and there are no firewall problems or other blockages is that someone may be in a situation in which listening and watching work fine but speaking (whether on video or not) may be impossible or inappropriate. If there is an alternative to typing a question under those circumstances, I don't know what it is. (4) There would be an advantage to everyone in the room being logged into Meetecho, particularly because it would allow using Meetecho to manage those who want to speak in a single, coherent, queue rather than relying on chairs to manage a remote queue and multiple mic lines and have their actions/ decisions be fair and perceived as fair. However, even if there are no bandwidth or performance issues, there is value in knowing (even crudely) who is f2f in the room and who isn't. There are also real advantages to using a real mic line for in-room participants even if Meetecho (or something else automated) is managing the queues. So, if we want to go down the path of many or most in-room participants running Meetecho, we may need to rethink the sign-in process and interfaces a bit. (5) To turn around something that has been mentioned, the requirement that someone in the mic line clearly identify themselves seems to be violated more often that not. It isn't just the people who don't mention their names, or don't mention them after the first interaction in a back-and-forth exchange (I'm sensitive to the issue and have still been guilty), but the ones who mumble their names or speak them too quickly, The latter, may be worse, and the definition of "quickly", may be more difficult if the speaking style or accent of the speaker differences from that of the listener. For experienced IETF participants, being able to see the speaker in combination with hearing whatever is said may help, but that does relative newcomers no good at all and doesn't help remote participants more generally if the camera is not pointed at the speaker. Having someone on Jabber recording who is speaking can be of huge help with this, at least unless they can't figure out who is speaking either. We have experimented with badge readers and other ways to make the speaker identification process more automatic and more reliable, but gave up on it, at least partially out of privacy concerns. Unless we can all get a lot better at identifying ourselves and doing so slowly, carefully, and clearly, maybe we need to review those decisions as tradeoffs between privacy and requirements for openness and transparency about who is making Contributions and influencing decisions. There is probably more, but I just reached my self-imposed length limit for this note :-( thanks, john
