TRAM - TURN Revised and Modernized IETF 90, Toronto (Canada) Wednesday, July 23, 2014 Chairs: Simon Perreault and Gonzalo Camarillo Note taker: Pål-Erik Martinsen * Chairs update The chairs gave a status update. They asked for volunteers to edit the STUN-bis and TURN-bis drafts. Gonzalo Salgueiro and Marc Petit-Huguenin volunteered to edit STUN-bis, and Tirumaleswar Reddy volunteered to edit the TURN-bis spec. Alan Johnston also volunteered to edit the TURN-bis spec right after this session. The chairs explained that the WG will not be taking on any more WG items beyond potentially draft-patil-tram-turn-serv-disc and draft-patil-tram-alpn-00, which are going to be discussed in this session, until there is progress on STUN-bis and TURN-bis. That is, until known errata have been incorporated into revised versions of the specsa and issues in the issue tracker are being actively worked on. * Multiplexing Scheme Updates for SRTP Extension for DTLS draft-petithuguenin-avtcore-rfc5764-mux-fixes-00 Presenter: Marc Petit-Huguenin This presentation was a heads up for the working group. The plan is to take this issue to the AVTCORE WG and work on it there. * TURN Extension for Third Party Authorization draft-ietf-tram-turn-third-party-authz-00 Presenter: Tirumaleswar Reddy Brandon Williams had some comments on the draft. He will follow up on the mailing list with a suggestion so that discussions can continue on the mailing list. Tiru has already contacted the OAuth experts to get the draft reviewed. The chairs will help find a reviewer if that ends up being difficult. * An Origin Attribute for the STUN Protocol draft-ietf-tram-stun-origin-00 Presenter: Alan Johnston It was mentioned that a typical origin with HTTPS could be something but with other protocols could be something else. So, using a MUST for the maximum length may not be a good idea. But the draft can give guidelines, though. There could be multiple origins in certain corner cases. But we could limit the number of origins or limit the attribute space. For next steps, we will get the draft reviewed by an HTTP expert. Alan has already talked with Martin Thompson and he is willing to review the next revision of the draft. * TURN Server Auto Discovery draft-patil-tram-turn-serv-disc-01 Presenter: Dan Wing The chairs need to figure out whether or not the HTTPBIS WG will work on standardizing proxy.pac. Also, we need to understand the relationship with W3C in this area. The chairs took a hum and there was a strong consensus to adopt this draft as a WG item. * Single SOcket Dual Allocation with TURN (SSODA) draft-martinsen-tram-ssoda-00 Presenter: Pål-Erik Martinsen Dan Wing advised not to limit the protocol to a particular set of address family pairs. Around 7 people in the room had read the draft and none of them thought this was a crazy idea. The authors will continue working on it (note that this draft is not a WG item). * Peer-specific Redirection for TURN draft-williams-peer-redirect-01 Presenter: Brandon Williams Ted Hardie asked for the draft to contain explanations on what kind of information is necessary for a TURN server farm operator to be able to perform redirections that actually improve connectivity over traditional mechanisms (i.e., anycast and split DNS). Pål-Erik Martinsen and Justin Uberti observed that a smarter ICE taking into account live measurements for better path selection would be good. Justin Uberti agreed with the presenter that a path using TURN can be faster than a path using direct connectivity. There were discussions on several topics. The authors will continue working on it (note that this draft is not a WG item). * A Bandwidth Attribute for TURN draft-thomson-tram-turn-bandwidth-01 Presenter: Alan Johnston Improvements to the draft were suggested. Among others: - Asymmetric bandwidth specification - Asynchronous feedback from the server upon changing network conditions using STUN Indications There were discussions on several topics. The authors will continue working on it (note that this draft is not a WG item). * Mobility with TURN draft-wing-tram-turn-mobility-00 Presenter: Pål-Erik Martinsen Around 10 people in the room had read the draft. There was a question about what happens when there are packet losses, especially when responses are lost and endpoints end up without a valid mobility ticket. The authors will look into how to resolve this. The authors will continue working on this draft (note that this draft is not a WG item). * Using DANE to validate TLS certificates for STUN draft-petithuguenin-tram-stun-dane-00 Presenter: Marc Petit-Huguenin Around 3 people in the room had read the draft and none of them thought this was a crazy idea. In fact, they thought this was a useful idea. The authors will continue working on it (note that this draft is not a WG item). * Application Layer Protocol Negotiation (ALPN) for STUN draft-patil-tram-alpn-00 Presenter: Gonzalo Salgueiro The chairs took a hum and there was a strong consensus to adopt this draft as a WG item. The meeting ended.