IETF 111 July 29, 2021
Original BOF was March 2011 and we are not done yet.
I-D:
https://datatracker.ietf.org/doc/draft-uberti-rtcweb-rfc8829bis/
Repo:
https://github.com/rtcweb-wg/jsep/pull/1017
Justin presented.
Ekr - add that the correct behavior is what was max-bundle.
Justin will add a bit of that text which clarifies prefered behavior to introduction section.
On the Issue with 3PCC:
Roman would like to see us to add a note that says it works this way and 3PCC implementation need to be aware but no normative changes to how the draft works. No objections to adding this.
Sean Turner - are we done? No issues raised.
Murray - There’s the examples document - it’s dead. Do we want to reconsider? Not invested in it.
Cullen - there’s useful info for implementers and we were going to pub, but don’t know what we were waiting for. I prefer to pub.
Sean - fine to run it through this WG.
Murray - I’ll check the charter.
Sean - we’ll check with Suhas.
Ted - we’ll take it to the list. The review of examples is difficult. It may have stalled on reviews.
Please send a note to chairs that you will review. We’ll ask Suhas if it’s ready.
Murray - if there’s not enough energy that’s ok too.
ekr - fine to keep as an I-D, and not advance - lot of work, there are defects in the doc, world has gone on.
Nils - I tried to review and ran out of steam. Devide and conquer I guess? One person to review would take a lot of time.
Sean - fine to split it up.
Cullen - we split it up last time, got good reviews. Let’s review the sections we didn’t get reviews on the last time. But asking for reviews and not publishing doesn’t encourage people to review.
Justin - machine generation found issues we thought were correct. Unless machine checked, there will be errors.
ekr - who is advocating to take it to RFC?
Bernard - 2nd Justin’s concerns. SDP has changed. Likely an enormous amt work required here. Not saying it’s not worth doing.
Justin - python based generator for JSEP, maybe could adapt that for this effort. Offering to help.
Bernard - reviewing the generator would be easier than the examples.
Nils - the draft abbreviates the SDP in the latter sections. Automating it will change the examples.
Jonathan Lennox - agree with ekr. Don’t think it’s worth the effort.
Cullen - everytime I have to debug webrtc, I have to look at SDP.
Timothy - audience would be implementers of the stack. You can debug SDP with browser tools. But it’s a small audience
Sean - we can solicit reviews and then hand it back to you, Murray.
Ted - first check into Suhas. If he doesn’t have energy, we should take that signal. After talking to him, the chairs will talk to Murray about AD-sponsored. Anyone object?
(no objections)
Ted - any other topic for the WG?
(none raised)