Thursday, July 21, 2016< ^ >
Dave Thaler has set the subject to: Transport Services WG
Room Configuration
Room Occupants

[16:23:57] Meetecho joins the room
[16:25:06] Simon Pietro Romano joins the room
[16:25:07] George Tsolis joins the room
[16:25:07] George Tsolis leaves the room
[16:27:23] Stein Gjessing joins the room
[16:28:08] John Leslie joins the room
[16:28:27] theresa joins the room
[16:29:07] theresa leaves the room
[16:29:08] Theresa Enghardt joins the room
[16:35:47] csp joins the room
[16:36:12] <csp> TAPS @ IETF 96
[16:36:19] <csp> WG status slides
[16:36:41] tale joins the room
[16:38:34] <csp> Slide: “Need for northbound info”
[16:38:55] <csp> Is there anyone remote? Is the AV working?
[16:40:49] George Tsolis joins the room
[16:41:25] Toru Asahina joins the room
[16:43:19] <csp> Update on draft-ietf-taps-transport-usage – Naeem Khademi presenting
[16:45:21] Kyle Rose joins the room
[16:47:45] <csp> Features of UDP and UDP-lite – Gorry Fairhurst presenting
[16:48:01] George Tsolis leaves the room
[16:51:54] <csp> Do we merge with draft-ietf-taps-transports-usage? No
[16:52:05] <csp> Update on draft-gjessing-taps-minset-02 - Michael Welz presenting
[16:56:39] <csp> Q: does “automatable” prohibit the application from controlling the features? (Aaron Falk) no
[16:57:26] <csp> Q: is this draft done? No. Updates coming.
[16:58:27] <csp> “On the cost of using happy eyeballs…” - Anna Brunstrom presenting
[16:59:53] <csp> Slides based on IRTF ANRW paper:
[17:08:51] <csp> Q (Tommy Polly, Apple): how parallel are the connection requests - simultaneous? Yes.
[17:09:43] <csp> Their (Apple’s) experience with IPv4/IPv6 Happy Eyeballs shows that biasing in favour of one protocol can significantly reduce costs. Can they do something similar for transport protocols?
[17:10:37] <csp> Anna has a-priori preference, but no learning and caching
[17:11:38] <csp> Q (Stephen Strowes, Yahoo): like the idea. how does happy eyeballs at the transport interplay with happy eyeballs at the network layer?
[17:12:02] <csp> How does it relate to ICE?
[17:12:23] <csp> Anna: Not studied
[17:12:38] <csp> Tommy (Apple): they have some experience. Needs to be done with care.
[17:14:26] <csp> Q (varun singh): ICE wg has related results; cullen jennings has data. they see 30+ candidates for ipv4/v6 today. this will make much worse.
[17:15:05] <csp> varun: congestion control for this is hard
[17:16:38] <csp> Colin Perkins: this happy eyeballs is essentially ICE. scope for coordination
[17:16:56] <csp> Michael Tuxen: ???
[17:17:29] <csp> “Post Sockets: A Generic API for…” - Brian Trammell presenting
[17:24:50] <csp> Colin Perkins: like this. add deadlines.
[17:25:27] <csp> Colin Perkins: related - the last presentation coming in the session; also, Clark and Tennenhouse paper on ALF (SIGCOMM, ’85?)
[17:25:48] <csp> Tommy: like
[17:26:00] <csp> “Socket Intents” - Philipp Tiesel presenting
[17:31:16] <csp> “Implementing real-time transport services over an ossified network” - Stephen McQuistin presenting
[17:31:24] Toru Asahina leaves the room
[17:31:32] <csp> Presentation based on
[17:40:18] <csp> Michael Tuexen: PR-SCTP does much of this
[17:41:01] <csp> Brian Trammel: is a little sceptical of the inconsistent retransmissions; would like to see larger scale measurements.
[17:41:21] <csp> End of WG session
[17:41:33] Kyle Rose leaves the room
[17:41:51] csp leaves the room
[17:41:54] John Leslie leaves the room
[17:42:10] Simon Pietro Romano leaves the room
[17:42:27] Stein Gjessing leaves the room
[17:43:47] Meetecho leaves the room
[17:46:49] Kyle Rose joins the room
[17:47:55] tale leaves the room: Disconnected: closed
[17:57:24] Theresa Enghardt leaves the room
[19:38:48] tale joins the room
[20:03:13] Kyle Rose leaves the room
[21:00:23] tenghardt joins the room
[21:00:47] tenghardt leaves the room
[21:59:26] csp joins the room
[21:59:51] csp leaves the room
Powered by ejabberd - robust, scalable and extensible XMPP server Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!