Hi.
I agree with Kathleen (and one or two others so far) and would
like to try to fork this discussion a bit.
Putting the question of assessment criteria to one side (those
who want to continue to refine those criteria should certainly
do so), the mere fact that such criteria are needed implies that
there is some reasonable possibility that we will be doing IETF
108 online. (The opinion expressed by several people on-list
that a remote IETF 108 is inevitable is not relevant to this
note.)
So can we devote some time and energy to figuring out how we are
going to do that meeting (and/or 109) online if we need to and
what properties are important? While I think IETF 107 worked
well given the short notice and circumstances, I don't think it
should automatically set the pattern for what we do with non-f2f
meetings in the future (for me, reading through the survey
results reinforces that view). I suggest it actually should not
do that.
Others may not agree, but I think two of the IETF's main
strengths and contributors to our success, are:
(1) We do, or used to do, a good deal of technical level
cross-protocol and cross-area review and interaction. At least
IMO, a great deal of that came from people wandering into f2f WG
sessions, sometimes out of curiosity, dragged by others, or for
other reasons, on an "I'm here anyway and not very busy, so
let's see what is going on" basis. I can remember times when WG
Chairs or ADs approached people at f2f meetings who were not
involved in the work of a particular WG but who appeared to
possibly have relevant expertise and asked them to sit in. No
matter what technology we use, it is not clear to me how to
replicate the useful effects of that kind of review and
interaction with online meetings, at least without thinking
things through and making specific plans. Whatever we might
think about it otherwise, it seems clear to me that an IETF-wide
Jabber "hangout" may be the solution to some other problems, but
not to this one. The discussion of the effectiveness of reviews
from various review teams and directorates that we had a while
back may also be interesting wrt this topic, but the mere fact
that those reviews are rarely initiated before IETF Last Call
starts puts them into a somewhat different category.
(2) We also used to have firm rules that all of our real work
was done on mailing lists and Internet-Drafts -- no f2f (with or
without subsequent summaries to mailing lists), trackers or
github-type arrangements as anything but aids to keeping
discussions organized. That model eliminated disadvantages due
to choices of time zones and made it fairly easy for someone who
was willing to sit down and do several hour's of reading to
understand the status of a particular work item and then
participate. At least some people here can remember when the
IESG aggressively pushed back on interim meetings (f2f or
online), required Ad authorization for each one, banned them
from occurring instead of meetings during regular IETF meetings,
only very rarely allowed more than one or two of them for a
given WG between IETF and asked tough questions about why the
work could not be done on email. While that rule about the use
of email is supposedly still in place, my (entirely anecdotal
and possibly incorrect) observation is that it has eroded
considerably. The recent near-epidemic of some WGs scheduling
multiple interim meetings may be a bad sign, at least in the
sense that it requires much more effort and scheduling than for
someone who is merely curious (but might end up contributing) to
read a draft or two, look through the mailing list archives, and
thereby catch up.
Now, it may be that I'm just an old guy suffering from a
hopeless case of nostalgia for the way I'd like to pretend the
IETF used to be. But, if either of the above are still
relevant, the time to have a discussion about how we do online
IETF meetings while preserving as much of the advantages of
those principles and mechanisms as possible should probably be a
topic of discussion RSN (if not sooner).
Along the same lines, one of the things that caused IETF 107 to
work was dropping all of the scheduled WG meetings, retaining
(approximately) only the plenary, some BOFs, and some area
DISPATCH-like meetings. Those were held more or less on
Vancouver time, presumably because IETF 107 was supposed to be
in Vancouver and that preserved the time zone properties of the
venue rotation model even if not the geographic properties
(maybe I'm wrong about that and the reasons were something else
entirely -- I don't recall the community being told).
Considering the discussions of a few months ago (repeated in
comments in the survey) about working in a different time zone
while being at home, it seems to me that there are arguments for
holding enough meetings during IETF week that many (or at least
some) of us can block out the whole week and just make the shift
(with jet lag without leaving home before and after) as well as
arguments for spreading the WG meetings out (ideally on
schedules that work for everyone who might be interested in a
given WG's work whether they have been, or expect to be, active
on not). What do we want to do about that, or will that be
another top-down decision and announcement? And, if IETF 108
is online, will it be scheduled around Madrid time in deference
to the former location or on some other schedule? Should we
adopt the principle that some WGs seem to be doing and rotate
times of the sessions? Specifically, should the plenary be
repeated three or four times for convenience in different time
zones? Or should most of it be done once, made available on
YouTube or equivalent immediately, and then question/microphone
periods repeated at several different times?
I don't pretend to have answers to those questions (or, in most
cases, even preferences) but, unless the community wants to
delegate decisions about all of them to the management team and
its ability to consult anyone they feel like and not anyone else
(and, in the process, agrees that the team the IESG pulled
together to make decisions about IETF 107 has the right
composition going forward), it seems to me this is the right
time to start asking the questions and discussing possible
answers. Others might, but I don't really care whether that
process is driven by the IESG, by Jay, or from the mailing list,
but I think it is important that it start if there is even the
vaguest of possibilities that either IETF 108 or 109 (or both)
will end up not being f2f.
thanks,
john
p.s. One of the other things that was apparently decided quickly
and quietly before IETF 107 was the switch from Meetecho to
WebEx for the plenary, BOFs and WG-type sessions. It seems to
me that each has advantages. Meetecho definitely had (and
probably has) a loyal following among at least some long-time
IETF remote participants and, while it has idiosyncrasies of its
own, we engineered around some of the glitches that showed up
with WebEx during the meeting week some years ago. I don't know
it is scales better than WebEx to a very large number of remote
participants who really intend to participate, but my impression
from IETF 107 (and some survey comments) was that WebEx didn't
scale perfectly either. So, while I don't think it makes sense
to try to make that decision on this mailing list, I hope
whoever makes it for future meetings will at least announce and
explain their reasoning to the community.
--On Sunday, April 19, 2020 11:37 -0400 Kathleen Moriarty
<kathleen.moriarty.ietf@xxxxxxxxx> wrote:
>...
> It would be nice to move on and figure out how to best run
> virtual meetings and improve upon learnings as opposed to
> having this discussions. Let's spend our time well.
>
> Best regards,
> Kathleen
