IETF
rmcat@jabber.ietf.org
Thursday, August 1, 2013< ^ >
rscheff has set the subject to: bloated buffers
Room Configuration
Room Occupants

GMT+0
[10:46:26] jgunn joins the room
[10:55:24] jgunn leaves the room
[10:57:16] jlcJohn joins the room
[11:01:02] Lars joins the room
[11:03:16] britram joins the room
[11:03:38] Olle E. Johansson joins the room
[11:03:43] wolfgang.beck01@gmail.com joins the room
[11:04:07] <Lars> i can relay questions from the jabber room
[11:04:17] Xavier Marjou joins the room
[11:05:56] <Olle E. Johansson> Ted H speaking
[11:06:02] <Olle E. Johansson> RTCweb chair
[11:06:19] jesup joins the room
[11:07:30] jgunn joins the room
[11:07:32] <Olle E. Johansson> Any remote participants?
[11:07:36] <jesup> yes
[11:07:40] <jgunn> yes
[11:07:52] <Olle E. Johansson> Any jabber scribe?
[11:07:54] <jlcJohn> trying to listen, anyway...
[11:08:03] tomkrist joins the room
[11:08:29] <jlcJohn> Lars said he could relay...
[11:08:33] <Olle E. Johansson> David black
[11:08:45] <Lars> i can relay one-way, form the jabber room to the meeting room
[11:09:06] <Olle E. Johansson> Colin Perkins in queue
[11:09:32] <Olle E. Johansson> Fluffy = dr Cullen Jennings co-chair of rtcweb
[11:09:46] <Olle E. Johansson> zahed sarker talking
[11:10:06] <Olle E. Johansson> Cullen Jennings talking
[11:10:25] <Olle E. Johansson> "some work needs to happen here okey" - Cullen
[11:11:03] <Olle E. Johansson> Lars speaking as chair
[11:11:31] <Olle E. Johansson> Colin Perkins
[11:11:50] Jonathan Lennox joins the room
[11:12:20] <Olle E. Johansson> MIchael <?> speaking
[11:12:29] <jesup> tuexen
[11:13:20] <Olle E. Johansson> Mirja speaking
[11:14:15] <jesup> I certainly hope they will at least coordinate instead of fighting, which touches on other congestion coordination thoughts
[11:14:30] <Olle E. Johansson> Lars: Please ask speakers to wear their badges in clear sight to make it easier for scribes to parse
[11:14:44] <Lars> ack
[11:14:51] <Olle E. Johansson> Michael A. Ramalho presenting
[11:15:15] <Olle E. Johansson> Slide 2 - RMCAT evaluation criteria design outline
[11:15:38] <Olle E. Johansson> Slide 3: RMCAT design team background
[11:16:30] <Olle E. Johansson> Focus on congestion control part today
[11:16:52] <Olle E. Johansson> Slide 4: RMCAT Protocol transport design
[11:17:00] <Olle E. Johansson> Oh, how I miss meetecho :-)
[11:18:44] <jesup> :-/  Yeah.
[11:18:51] <Olle E. Johansson> If you need relay by Lars, please prefix your statement with "mic:"
[11:19:13] <Lars> or mention my name, since my jabber client highlights that nicely
[11:19:22] <Olle E. Johansson> Slide 5: RMCAT evaluation test topology
[11:20:02] gmaxwell joins the room
[11:20:35] <Olle E. Johansson> Colin Perkins waiting at mic
[11:20:41] <Olle E. Johansson> Colin talking
[11:21:20] <jesup> mic: (slide 4) - we want to live with other delay flows except maybe if we're not in a Codel/etc environment (just for clarification)
[11:21:20] <Olle E. Johansson> The room is quite full
[11:22:06] <Olle E. Johansson> Slide 6: RMCat design team initial experiments
[11:22:15] <Olle E. Johansson> lars: relay request above!
[11:23:56] <Olle E. Johansson> Jana ? talking
[11:24:09] <britram> yes, Jana.
[11:24:23] <Olle E. Johansson> Zahed talking
[11:25:13] <Olle E. Johansson> Colin Perkins talking
[11:25:34] <Olle E. Johansson> Comments on slide 6
[11:26:02] <Olle E. Johansson> As a Jabber scribe I would prefer having slide numbers… ;-)
[11:26:50] Piers O'Hanlon joins the room
[11:28:25] <Olle E. Johansson> Mo talking
[11:29:02] <Olle E. Johansson> Colin responding
[11:29:55] <Olle E. Johansson> Toerless Eckert talking
[11:30:31] <Olle E. Johansson> Colin responding
[11:33:34] <Olle E. Johansson> Still on slide 6
[11:34:12] <Olle E. Johansson> Lars asking
[11:34:20] <Olle E. Johansson> …as individual I guess
[11:34:23] <Olle E. Johansson> :-)
[11:35:21] <Olle E. Johansson> Mirja Kuehlewind talking
[11:36:02] <Olle E. Johansson> Yes, there are two bullet #6
[11:36:06] <Olle E. Johansson> Zahed talking
[11:36:50] Makkes joins the room
[11:36:53] <Olle E. Johansson> Slide #7: Open Issues, discussions & Future challenges
[11:38:04] <Olle E. Johansson> "Please join the design team" - Michael
[11:41:23] <Olle E. Johansson> Zahed talking
[11:41:48] <Olle E. Johansson> David black talking
[11:42:02] <Olle E. Johansson> ? speaking
[11:42:24] <jesup> mic: Bursty flows are a common case where browsing is happening on the same host or on another device behind the same bottleneck (wifi router, etc), and should be used as a model for competing bursty traffic (at least to the first approximation)
[11:42:31] <Lars> ? = rüdiger geib
[11:42:35] <Olle E. Johansson> Next speaker
[11:42:38] <jesup> lars: ^
[11:42:45] <britram> lars
[11:42:54] <Olle E. Johansson> Varun Singh
[11:43:04] <Olle E. Johansson> draft-singh-rmcat-cc-eval <http://tools.ietf.org/html/draft-singh-rmcat-cc-eval>
[11:43:16] <Olle E. Johansson> Slide 2: Changes since -02
[11:43:22] <jesup> lars: never mind, not important
[11:43:48] <Olle E. Johansson> ok
[11:44:13] <Lars> sorry for missing the remote comments. maybe it would be better if someone else relayed?
[11:44:21] <Lars> i suck at multitasking
[11:44:45] <Olle E. Johansson> :-)
[11:44:53] <Olle E. Johansson> That's why I like someone else doing it too :-)
[11:45:23] <Olle E. Johansson> We will solve this, so keep comments comming while we organize ourselves in the room...
[11:45:44] <Olle E. Johansson> Slide 3: Unfairness Metric
[11:46:36] <Olle E. Johansson> Jonathan: Could you help with relay services?
[11:46:57] <Olle E. Johansson> zahed talking
[11:47:24] <Olle E. Johansson> Colin Perkins waiting in line
[11:47:33] <Olle E. Johansson> Zahed talking again
[11:47:56] <Olle E. Johansson> Colin Perkins talking
[11:48:20] <Olle E. Johansson> Lars talking
[11:48:40] <Olle E. Johansson> ? talking
[11:48:45] <Lars> michael welzl
[11:48:48] <britram> Michael Welzl
[11:48:55] <Olle E. Johansson> Thanks
[11:49:19] <Olle E. Johansson> Slide 4: Metrics: Open issues
[11:49:43] <Olle E. Johansson> Zahed talking
[11:51:42] gmaxwell leaves the room
[11:51:52] <Olle E. Johansson> Mo talking
[11:52:05] <Olle E. Johansson> Lars talking
[11:52:12] wolfgang.beck01@gmail.com leaves the room
[11:53:20] <Olle E. Johansson> Colin Perkins talking
[11:54:16] <Olle E. Johansson> zahed talking
[11:54:37] <Olle E. Johansson> Xiaqoing Zhu waiting in line, as in Ken Carlberg
[11:55:28] <Olle E. Johansson> Xiaoqing Zhu talking
[11:56:22] <Olle E. Johansson> Ken Carlberg
[11:56:27] <Olle E. Johansson> TAlking about MOS Scores
[11:56:38] <Olle E. Johansson> MOS Scores applies to audio
[11:56:51] <Olle E. Johansson> MOS for video exists
[11:57:14] <Olle E. Johansson> Mirja and Lars speaking
[11:57:53] <Olle E. Johansson> Slide 5: Evaluation parameters
[11:58:50] <Olle E. Johansson> Xiaoqing Zhu talking
[11:59:43] <Olle E. Johansson> Zahed talking
[12:00:16] <Olle E. Johansson> Lars talking
[12:00:23] <jesup> mic: TCP cross-traffic will need to include long term flows, and also short-term/bursty queues.  The requirements document says we need to work well with AQM queues - algorithms for delay-based could easily fail badly in an AQM situation
[12:00:35] <jesup> lars: ^
[12:00:49] <Olle E. Johansson> Slide 6: List of traffic flows
[12:01:04] hardie@jabber.psg.com joins the room
[12:01:15] wolfgang.beck01@gmail.com joins the room
[12:01:43] <Olle E. Johansson> Lars talking (for himself)
[12:02:37] <Olle E. Johansson> Slide 7: Parameters open issues
[12:02:58] scimbe@draugr.de joins the room
[12:03:06] hardie@jabber.psg.com leaves the room
[12:03:28] Olle E. Johansson has set the subject to: rmcat - Agenda http://tools.ietf.org/wg/rmcat/agenda
[12:03:36] <jesup> mic: disagree with lars: we do need to consider delay-based flows other than the selected one, since people use algorithms on the net which aren't appoved/specced by the IETF (frightening, I know ;-)
[12:03:40] <jesup> lars: 6
[12:04:14] <jesup> It doesn't mean we have to work well with them, but we need to realize we'll see them
[12:04:18] <jesup> mic: ^
[12:04:23] <Olle E. Johansson> done
[12:04:33] <Olle E. Johansson> Easier for me to disagree on your behalf :-)
[12:04:47] <Olle E. Johansson> Lars answering
[12:04:49] <Lars> thanks
[12:04:50] <Olle E. Johansson> Ken carlberg talking
[12:06:09] <Olle E. Johansson> Slide 8: Next steps
[12:06:25] <Olle E. Johansson> Lars talking
[12:07:05] <Olle E. Johansson> Sorry, missed latest mic relay - still want it?
[12:07:09] <Olle E. Johansson> Mirja talking
[12:07:14] Xavier Marjou leaves the room
[12:07:19] <Olle E. Johansson> @jesup ^
[12:08:14] <Olle E. Johansson> Lars talking
[12:08:45] <Olle E. Johansson> "I think it was great and I think we're so not done yet" - Lars (as chair)
[12:08:51] <Piers O'Hanlon> Would it make sense to provide ns2/3 baseline of the parameters?
[12:09:07] <Olle E. Johansson> Mirja talking
[12:09:18] <jesup> read
[12:09:30] <Olle E. Johansson> Xiaoqing talking
[12:09:36] gmaxwell joins the room
[12:09:53] <Olle E. Johansson> Suggesting we split the draft
[12:09:54] <jesup> and I support adopting it as a WG item (as either 1 or two docs)
[12:09:57] <Olle E. Johansson> Lars responding
[12:10:21] hardie@jabber.psg.com joins the room
[12:10:32] <Olle E. Johansson> Next presentation - we are 10 minutes late
[12:10:32] <Olle E. Johansson> Safiqul Islam
draft-welzl-rmcat-coupled-cc <http://tools.ietf.org/html/draft-welzl-rmcat-coupled-cc> (
[12:10:40] <Olle E. Johansson> Presenter: Safiqul Islam
[12:10:54] <Olle E. Johansson> Slide 1: intro
[12:11:41] <Olle E. Johansson> Slide 2: A not about evaluations
[12:11:51] <Olle E. Johansson> slide 3: Why do we need coupled cc?
[12:12:23] <Olle E. Johansson> Slide 4: How to solve this
[12:13:05] <Olle E. Johansson> Slide 5: Flow state exchange
[12:14:15] <Olle E. Johansson> Slide 6: FSE
[12:15:19] <Olle E. Johansson> The room is full, lots of interest in this topic
[12:15:58] <Olle E. Johansson> slide 7: FSE- how it works
[12:16:01] Xavier Marjou joins the room
[12:16:53] Makkes leaves the room
[12:16:56] Makkes joins the room
[12:17:00] <Olle E. Johansson> Slide 10: FSE - how it works (part 2)
[12:17:06] <jesup> Lost audio almost - getting picked up by some other mic far away from speaker
[12:17:13] <Olle E. Johansson> Slide 11: FSE - how it works
[12:17:26] <Olle E. Johansson> Very small changes between the slides
[12:17:33] <britram> speaker still amplified in the room
[12:18:14] <jesup> hmmm
[12:18:15] <Olle E. Johansson> Lars: We need to adjust the mike!
[12:18:37] <jgunn> audio OK for me
[12:18:37] <jesup> Freaky
[12:18:45] <jesup> I reloaded and the audio is better
[12:18:48] <Olle E. Johansson> Sorry - there's a lot of slide magic here I can't follow the changes in the animation
[12:18:49] <scimbe@draugr.de> audio ok for me too
[12:18:52] <jesup> maybe a server issue
[12:19:01] <Olle E. Johansson> Ok, then we don't do anything now. Thanks for the feedback
[12:19:19] hardie@jabber.psg.com leaves the room
[12:19:41] <Olle E. Johansson> IETF slides like this does NOT help scribes and remote participation. We need slide numbers
[12:20:31] hardie@jabber.psg.com joins the room
[12:20:42] <Olle E. Johansson> Next slide: Simulation results
[12:21:06] <Olle E. Johansson> Let's call it slide 42 for simplicity
[12:21:21] <Olle E. Johansson> Lars asks for simplification
[12:21:53] <Olle E. Johansson> Michael at mike
[12:22:24] <Olle E. Johansson> Zahed (again!)
[12:22:53] <jesup> I've been unable to view the slides with LibreOffice - trying OpenOffice but doubt it will work
[12:23:07] <Olle E. Johansson> Ouch
[12:23:20] <Olle E. Johansson> Michael back
[12:23:50] <Olle E. Johansson> Mirja
[12:23:56] <Olle E. Johansson> Michael talking
[12:24:28] <Olle E. Johansson> zahed
[12:24:57] <Olle E. Johansson> MIchael and Zahed speaking
[12:25:00] <jesup> Wow, for once OpenOffice works better and opened it
[12:25:13] <Olle E. Johansson> Roberto waiting
[12:25:30] <Olle E. Johansson> zahed
[12:25:46] <Olle E. Johansson> Slide 43: Priority
[12:25:59] <jesup> but all the graphics/images are missing
[12:26:07] <Olle E. Johansson> Arrrggghhhh.
[12:26:15] <Olle E. Johansson> Don't we have a pdf somewhere? Chairs?
[12:26:26] <Olle E. Johansson> Slide 44: Fairness
[12:27:19] <Olle E. Johansson> Slide 45: fairness
[12:27:29] <jesup> 44?  I see a total of 34 slide; seems like we're around 28?
[12:27:40] <Olle E. Johansson> I lost track and restarted on 42. Sorry.
[12:27:45] <Olle E. Johansson> I don't have the slides.
[12:27:45] <britram> there's a whole lot of builds
[12:27:56] <Olle E. Johansson> What's the current number for you?
[12:28:02] <britram> the slide number in the lower right corner is 14
[12:28:12] <Olle E. Johansson> Ahh! I missed that.
[12:28:21] <Olle E. Johansson> Slide #14 is the current one - "fairness"
[12:28:28] <britram> it's the 0.5-point text in the lower right. :)
[12:28:38] <Olle E. Johansson> Right in front of my nose.
[12:29:00] <Olle E. Johansson> Michael talking
[12:29:22] <jesup> aha.  Though these are all images, so I see nothing.  
[12:29:32] <britram> stand by working on a savedown
[12:29:57] <Olle E. Johansson> Zahed, Michael sharing microphone
[12:30:12] <Olle E. Johansson> Lars talking
[12:31:24] <Olle E. Johansson> Varun waiting by the mic
[12:31:36] <Olle E. Johansson> Michael and Lars talking
[12:31:42] <Olle E. Johansson> Varun talking
[12:31:58] <Olle E. Johansson> Mirja talking
[12:33:23] wolfgang.beck01@gmail.com leaves the room
[12:33:58] <Olle E. Johansson> Varun, Michael and speaker discussing
[12:34:37] <Olle E. Johansson> Mirja speaking as individual
[12:35:08] <Olle E. Johansson> Mo talking
[12:35:23] <Olle E. Johansson> Mo and Mirja discussing
[12:35:33] <Olle E. Johansson> Zahed talking
[12:36:13] Makkes leaves the room: Replaced by new connection
[12:36:14] Makkes joins the room
[12:36:29] Makkes leaves the room
[12:36:39] <Olle E. Johansson> Michael talking
[12:36:50] Makkes joins the room
[12:37:13] <britram> http://www.trammell.ch/ietf/slides-87-rmcat-2.pptx.pdf
[12:37:20] <jesup> britram: likely it's the "pptx" aspect.  I'm having pproblems with the NADA one as well - similar problems
[12:37:23] <Olle E. Johansson> Thanks britram!
[12:37:30] <britram> saved down to pdf in office for Mac.
[12:37:39] <jesup> thanks!  Can you save 3 too?
[12:38:01] <britram> at least the mac pdf renderer can see the slides.
[12:38:01] <Olle E. Johansson> Mirja talking
[12:38:01] <Olle E. Johansson> Michael answering
[12:38:01] <britram> aye
[12:38:02] wolfgang.beck01@gmail.com joins the room
[12:38:03] <britram> also taking notes so
[12:38:59] <Olle E. Johansson> Slide 16: Benefits from the non-greedy flows
[12:39:03] <jesup> -2 looks good in pdf, thanks
[12:39:19] <britram> downloading 3 now
[12:40:12] <Olle E. Johansson> Slide #18 Average queue length - 2 flows
[12:40:51] <Olle E. Johansson> Slide 20
[12:40:58] <Olle E. Johansson> Throughput for 2 flows
[12:41:22] <Olle E. Johansson> Slide #21 future plans
[12:42:02] <Olle E. Johansson> That's the last slide-----------
[12:42:19] <Olle E. Johansson> Mirja talking
[12:42:46] <Olle E. Johansson> Next presentation: Xiaoqing Zhu
[12:43:07] <Olle E. Johansson> Michael talking
[12:43:18] <Olle E. Johansson> Mo talking
[12:43:28] hardie@jabber.psg.com leaves the room
[12:43:55] <Olle E. Johansson> Lars speaking
[12:44:36] <Olle E. Johansson> "it seems to add complication at a time when that's the least thing we need" - Lars
[12:45:14] <Olle E. Johansson> Audio ok?
[12:45:23] <Olle E. Johansson> Slide #1 - Update on NADA
[12:45:29] <Xavier Marjou> ok
[12:45:44] <Olle E. Johansson> Slide #2 Agenda
[12:45:44] <Xavier Marjou> slides too http://tools.ietf.org/agenda/87/slides/slides-87-rmcat-3.pdf
[12:46:11] <jesup> thanks!
[12:46:12] <britram> http://www.trammell.ch/ietf/slides-87-rmcat-3.pptx.pdf
[12:46:17] <jgunn> Can't open this slide deck.
[12:46:26] <Olle E. Johansson> Slide #3 Recap: Design goals
[12:46:33] <britram> ah, oops. now there's two copies. :)
[12:46:42] <Olle E. Johansson> A fork, a fork.
[12:46:44] <Olle E. Johansson> Oh no!
[12:46:57] <jgunn> That helped.  Thanks
[12:48:07] <Olle E. Johansson> Slide #4 Recap Nada Receiver Behaviour
[12:49:57] <Olle E. Johansson> Slide #5: NADA sender behaviour
[12:51:48] <Olle E. Johansson> Animations on the slides again
[12:52:17] <Olle E. Johansson> Slide #6 implementation status
[12:53:12] <Olle E. Johansson> Slide #7: Simulation scenarios
[12:54:36] <Olle E. Johansson> Slide #8 A suite of queeue...
[12:56:08] <Olle E. Johansson> Sorry
[12:56:11] <Olle E. Johansson> We're at
[12:56:23] <Olle E. Johansson> Slide 10 Time-Varying link: Nada+red
[12:56:33] <Olle E. Johansson> (checked e-mail ;-) )
[12:56:59] wolfgang.beck01@gmail.com leaves the room
[12:57:32] <Olle E. Johansson> Slide #112: Interaction with various qeueing schemes
[12:57:36] <Olle E. Johansson> #12
[12:58:16] <Olle E. Johansson> Slide #13 Multiple competing NADA streams
[12:58:52] <Olle E. Johansson> Slide #14: Multiple competing NADA streams
[12:59:26] <Olle E. Johansson> Slide 15: NADA competing with TCP
[13:00:22] <Olle E. Johansson> Slide #16 Nada competing against TCP
[13:00:24] <Olle E. Johansson> Slide 17
[13:00:26] <Olle E. Johansson> Moving fast
[13:00:28] <Olle E. Johansson> Slide 18
[13:00:54] <Olle E. Johansson> Slide 19 summary
[13:01:25] <Olle E. Johansson> Xiaoqing seems to have EKR's baudrate now under stress from chairs. But not quite as fast.
[13:01:34] <jesup> Nice work
[13:01:49] <jesup> And kept understandable pronunciation :-)
[13:02:04] <Olle E. Johansson> Jim ? talking
[13:02:12] <Lars> gettys
[13:03:30] <jesup> mic: when competing against TCP with droptail, do you kinow why the first NADA flow doesn't change when a second was added?  Was there some floor that stopped it from driving into the ground?
[13:03:36] <jesup> lars: ^
[13:04:04] <Olle E. Johansson> ? speaking
[13:04:48] gmaxwell leaves the room
[13:05:00] <britram> that was rong pan
[13:05:13] <Olle E. Johansson> Logging out — -Thanks all!
[13:05:27] <Xavier Marjou> thank you too
[13:05:34] Olle E. Johansson leaves the room
[13:05:39] Xavier Marjou leaves the room
[13:05:45] Lars leaves the room
[13:05:45] Makkes leaves the room
[13:05:48] Jonathan Lennox leaves the room
[13:05:48] <jesup> thanks everyone
[13:05:55] <britram> ciao!
[13:06:04] Makkes joins the room
[13:06:16] Piers O'Hanlon leaves the room
[13:09:25] britram leaves the room
[13:11:29] jgunn leaves the room
[13:22:36] Olle E. Johansson joins the room
[13:23:01] Olle E. Johansson leaves the room
[13:25:03] Makkes leaves the room: offline
[13:32:20] Jonathan Lennox joins the room
[13:47:36] gmaxwell joins the room
[13:47:37] tomkrist leaves the room
[13:52:05] jesup joins the room
[13:52:23] tomkrist joins the room
[13:52:49] tomkrist leaves the room
[13:53:04] Jonathan Lennox leaves the room
[13:58:22] jesup leaves the room
[13:59:34] wolfgang.beck01@gmail.com joins the room
[13:59:36] britram joins the room
[13:59:42] wolfgang.beck01@gmail.com leaves the room
[14:01:00] jesup joins the room
[14:01:25] britram leaves the room
[14:01:28] britram joins the room
[14:07:23] jesup leaves the room
[14:14:48] britram leaves the room
[14:19:47] britram joins the room
[14:20:32] britram leaves the room
[14:20:35] britram joins the room
[14:52:57] britram leaves the room
[14:55:08] britram joins the room
[14:56:55] britram leaves the room
[14:56:58] britram joins the room
[14:57:40] britram leaves the room
[14:57:41] britram joins the room
[14:58:32] gmaxwell leaves the room
[15:00:10] britram joins the room
[15:00:35] britram leaves the room
[15:00:41] britram leaves the room
[15:11:41] jesup joins the room
[15:15:46] jesup leaves the room
[20:22:09] scimbe@draugr.de leaves the room
[22:04:03] jlcJohn joins the room
Powered by ejabberd Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!