# RTCWEB@IETF110 Chairs: Sean Turner, Ted Hardie Responsible AD: Murray Kucherawy Note Taker: Eric Rescorla ## Logistics ### Time FRIDAY, March 12, 2021 @ 1300-1500 UTC ### Meeting Information Meetecho (a/v and chat): https://meetings.conf.meetecho.com/ietf110/?group=rtcweb&short=&item=1 Audio (only): http://mp3.conf.meetecho.com/ietf110/rtcweb/1.m3u CodiMD: https://codimd.ietf.org/notes-ietf-110-rtcweb ## Agenda ### Administrivia Chair's Slides: https://datatracker.ietf.org/meeting/110/materials/slides-110-rtcweb-rtcweb-chairs-slides-ietf-110-01 ### "m=" re: bundle-only Justin Uberti's Slides: https://datatracker.ietf.org/meeting/110/materials/slides-110-rtcweb-aligning-jsep-bundle-00 #### Aligning JSEP+BUNDLE Resolution 1. 1. Switch BUNDLE to match JSEP [Christer volunteered to do it]. (14 hands in favor, none against) 3. Create a new "max-bundle-safe" [name TBD] with current "max-bundle-behavior". Document that "max-bundle" used to exist and now has unspecified and deprecated behavior. (14 hands in favor, none against) Separately, libwebrtc will try to migrate to specified behavior [non-IETF action] Need to discuss with ADs where bundle work happens. Plan is to re-rev JSEP in place rather than to do a patch draft. Christer will discuss the same thing for BUNDLE with the ADs and chairs. #### a=bundle-only for a m= line not in a BUNDLE group What happens if someone sends you a=bundle-only for an m= line not in a bundle group (can happen if an endpoint just copies attributes) Option 1: BUNDLE implementations must ignore a=bundle-only for m= lines not in a BUNDLE group Option 2: Generate an error (some debate about whether it actually is an error according to the current specs) Resolution: discuss on mmusic list and with ADs whether we should do this. ### NAT Slipsteam Harald Alvestrand's Slides: https://datatracker.ietf.org/meeting/110/materials/slides-110-rtcweb-nat-slipstream-00 Sense of the room is that this is not specific to RTCWEB and should be handled elsewhere. Should be discussed at DISPATCH.