Kent,
Excellent idea and brilliant summary !
I would add one more PRO (which one could think should be in IETF's best interest):
Make those who can not attend IETF meetings equal class citizens of IETF standardization work.
Today we see in vast majority attendees from vendors and providers of various Internet services who pay for its employees to promote their products and ideas or simply sponsor them to keep up with IETF work. It is virtually impossible for someone unknown to remotely propose an idea, run through IETF machine and get RFC out.
I think this is also one of the fundamental problems with limited innovation in networking space. The doors are generally closed. If you are a student and want to propose your brilliant P4 idea you are working on it, testing it and publish via github or docker hub. If you are an enterprise engineer it is really quite tough to sell value of IETF to anyone. Same for startups where they watch every dollar spent.
And as others said I do like to participate on site and meet folks, talk face to face, go for a beer etc .... but I also recognize that this is getting harder and harder for wider group of people year by year. And I do think we should really encourage new generations to get involved in Internet standards globally by making IETF work not only open and read-only accessible to anyone, but also allowing empowering anyone equally to propose ideas and standardize them.
Kind regards,
Robert.
Robert.
On Thu, Feb 27, 2020 at 6:26 PM Kent Watsen <kent+ietf@xxxxxxxxxx> wrote:
The premise of a few ongoing threads (and others over the years) is that there’s a need to preserve having synchronous communication amongst participants. Stated rationale include the ability to close issues more quickly to just getting to know each other.
Certainly knowing each other’s mannerisms is incredibly important, so as to have that understanding when conversation spills over to the list, but I question how often issues close more quickly, as all is to be closed on the list anyways, which is especially important to do when key participants don’t show, as effectively happens at every meeting already.
I propose moving IETF to pure-virtual *asynchronous* meetings as follows:
1) let each speaker give a time-bound virtual presentation including both “slides" and “facetime”, whereby “slides” may include other forms of screen share (e.g., to show a demo), and “facetime” is a videobox showing the presenter talking throughout the presentation (note that this is purposely audio+video, as opposed to audio-only, so as to help people to get to “know” the presenter).
1a) the presentation is as usual, a focus on the open issues, ideally a preface providing background info for those that aren’t following happenings as closely as others.
1b) the presentation should have "pause points" where feedback is being solicited. Depending on the nature of the presentation, there may be a pause point for each open issue, or perhaps just one after the entire presentation is complete. For example, adopted work will likely have a multiplicity of open issues, whereby unadopted work may not.
1c) the presentation must be reviewed by the chairs prior to publication, to ensure it meets IETF guidelines, stays within the allotted time, etc.
2) provide a bounded number of days for responses, responses to responses. etc. For instance, assuming an upper-limit of three-rounds for exchanges (until the issue has to be taken to the list), this limit may be 3 days..
2a) let each response be either audio-only, audio+video, or audio+video+screenshare. Text-only responses are not supported (and facetime encouraged!) so as to foster the "getting to know each other” aspect of the experience.
2b) anonymous responses are not permitted. IETF registration required (same as needed to use the MeetEcho “mic” today). This also enables private side-channels amongst participants (e.g., “you mentioned FOO, which I’m also interested in…”)
2c) with trepidation I’ll suggest also perhaps not permitting anonymous views (i.e. IETF registration required). This is in part for the chairs to understand who all viewed the presentation, especially the key participants, so as to gauge consensus. But this is also in part to aid the IESG in understand overall community interest in the WG and, perhaps, accrue a little revenue to cover the cost of the infrastructure needed to host the “meeting”.
2d) tooling would be needed to support threaded-responses and chair-driving humming on issues. Whilst registration would be required to hum, the hums can be anonymous (shielded from the chairs), as only the count totals (especially as compared to the virtual blue sheets) matter.
3) Miscellanea
3a) IETF Plenary, etc. would be run the same way. Perhaps with mandatory viewing prior to participants being able to view any WG “session”.
3b) Same for each WG, there could be a mandatory to view "Chairs Intro” before being able to view any of the other presentations.
3c) The idea is that the entire “meeting” would occur in a single week. Note that there is no issue with WG overlaps - all WG sessions could open at the same time starting Monday, with the response-cutoff being at the end of the week (skewed for timezones, so no one has to work over the weekend).
3d) Everything would be archived for posterity.
I'll stop here because, hopefully, the gist of the idea is clear and the message is already long, but let me throw out a quick scoring:
PROs:
1) The dependence on physical travel is removed.
2) The dependence on reliable broadband is removed.
3) Everyone can stay comfy in their local timezones.
4) Enables non-English native participants time to digest material.
5) Avoids having to scale infrastructure to support synchronous attendance.
6) Anything else?
CONs:
1) In-person communication benefits are lost.
2) Adhoc meetings (hallway huddles, meals, etc.) are not catered to.
3) Presents an opportunity for corporate gaming (e.g., “prepared" responses)
4) Ability to show off badge-stickers eliminated ;)
5) No ability to share gifts
6) Anything else?
