IETF
rmcat@jabber.ietf.org
Thursday, 2 August 2012< ^ >
Room Configuration

GMT+0
[19:10:08] Varun Singh joins the room
[19:52:03] Ted.h joins the room
[19:54:31] spencerdawkins joins the room
[19:58:57] JohnLeslie joins the room
[19:59:54] Cullen Jennings joins the room
[20:00:01] Michael Tüxen joins the room
[20:01:10] martin.stiemerling joins the room
[20:02:58] Ingemar Johansson joins the room
[20:04:03] Ingemar Johansson leaves the room
[20:04:08] Ingemar Johansson joins the room
[20:04:38] <Ingemar Johansson> Testing
[20:05:25] <Ted.h> Hi Ingemar
[20:05:29] <Ted.h> You are visible in chat
[20:05:42] wolfgang.beck01 joins the room
[20:05:47] wolfgang.beck01 leaves the room
[20:06:02] Wolfgang Beck joins the room
[20:06:04] <Ingemar Johansson> Thx
[20:07:03] <Michael Tüxen> Are slides available online?
[20:07:30] Jonathan Lennox joins the room
[20:08:19] Michael Richardson joins the room
[20:08:20] frodek joins the room
[20:09:05] sftcd joins the room
[20:09:18] <Michael Tüxen> Yes
[20:09:18] Martin Thomson joins the room
[20:09:19] <Ted.h> If you need something channeled to the room, please put MIC: in front, thanks
[20:09:25] Magnus joins the room
[20:11:26] tterribe joins the room
[20:11:49] tterribe leaves the room
[20:12:44] <Cullen Jennings> wow buffer bload
[20:13:40] xiaoqing.zhu joins the room
[20:14:18] <JohnLeslie> Are Harald's slides available anywhere?
[20:14:28] <Ted.h> Just went up on the proceedings site
[20:14:45] <JohnLeslie> thx
[20:15:23] <Ingemar Johansson> Which presentation
[20:15:58] <Ingemar Johansson> RMCAT problem statement ?
[20:16:06] <Jonathan Lennox> yes
[20:16:38] <Michael Tüxen> Can you provide a URL?
[20:16:51] <Ingemar Johansson> http://www.ietf.org/proceedings/84/slides/slides-84-rmcat-5.pdf
[20:17:06] <Michael Tüxen> Thanks
[20:17:26] <Varun Singh> thanks
[20:18:01] Dan York joins the room
[20:18:16] martin.thomson joins the room
[20:18:31] martin.thomson has set the subject to: bufferbloat
[20:18:45] Martin Thomson leaves the room
[20:19:24] Zahed Sarker joins the room
[20:21:03] mramalho@cisco.com joins the room
[20:21:33] Matt Zekauskas joins the room
[20:23:31] <Dan York> "It will take a little time to revolutionize the world." - Harald
[20:23:55] xiaoqing.zhu leaves the room
[20:24:45] rillian joins the room
[20:26:03] tsuichi joins the room
[20:26:06] <Cullen Jennings> less time with WebRTC, I mean twitter
[20:26:21] hta joins the room
[20:26:39] <Jonathan Lennox> Does someone have the URL to this slide deck?
[20:26:53] <martin.thomson> https://bufferbloat/
[20:27:18] <Varun Singh> http://www.iab.org/wp-content/IAB-uploads/2012/07/7-cc-workshop-summary-revised.pptx
[20:27:25] <Dan York> I don't see them on the materials page
[20:27:29] <Ted.h> Go here: https://datatracker.ietf.org/meeting/84/materials.html and look for the RMCAT section
[20:27:50] <Dan York> Don't see these, Ted
[20:28:22] <Ted.h> Yes, sorry. I thought the chair slides included these, but I was wrong.
[20:28:31] <Ted.h> But that's the general space for search
[20:29:00] <Dan York> Yep ... they are at the iab.org <http://iab.org> URL that was included above by "vr000m"
[20:29:10] <Dan York> We're on slide 4
[20:29:20] <Dan York> Anyone remote?
[20:29:32] <Varun Singh> Yes I am remote
[20:29:43] <Ted.h> Now slide 5
[20:30:06] <Dan York> Okay, if you need something relayed let us know
[20:30:32] <Dan York> (and maybe one of us will get up out of our chairs and relay your comment to the mic)
[20:30:49] <Ted.h> Hi Dan, I agreed to watch the room for jabber comments
[20:31:05] mreavy joins the room
[20:31:08] <Varun Singh> ok great.
[20:31:09] <Ted.h> So, MIC: at the front is the general way to distinguish those.
[20:31:22] <Dan York> Awesome, Ted!
[20:31:49] <Ted.h> Now on Jim's slides, slide 2
[20:31:55] <Ted.h> They are not numbered
[20:33:00] <Ted.h> Now on "Bufferbloat status"
[20:33:39] michael.welzl joins the room
[20:33:55] tomkrist joins the room
[20:34:45] Lars joins the room
[20:34:52] <Ted.h> Now on "no silver bullet" slide
[20:37:26] <Ted.h> AQM Algorithms slide
[20:38:30] xiaoqing.zhu joins the room
[20:39:14] <Ted.h> Now on "CoDel" slide
[20:39:23] spencerdawkins leaves the room
[20:40:05] tomkrist leaves the room
[20:40:08] tomkrist joins the room
[20:40:30] <Ted.h> Now on "Codel Status"
[20:41:53] cabo joins the room
[20:42:03] <Lars> so far, i have heard more on bufferbloat than on congestion control… how many more slides?
[20:42:12] <Ted.h> Now on "Think out of the Box"
[20:42:24] <michael.welzl> 2
[20:42:25] <Ted.h> Lars: please put "MIC" in front if you would like that relayed.
[20:42:31] <Lars> i'm in the room :-)
[20:42:34] <Lars> just venting
[20:42:42] <Cullen Jennings> @lars: on bufferbload ? or on Congestion ? prob zero to later
[20:43:01] <michael.welzl> the agenda says, until :45
[20:43:11] <Lars> but he started early
[20:43:15] EKR joins the room
[20:43:26] <EKR > Is there some relevance here?
[20:43:30] <michael.welzl> true. but not much - and we're well within time
[20:44:01] <hta> just don't ask questions, and it goes faster.
[20:44:17] <Ted.h> Home router disaster
[20:44:18] <EKR > Really?
[20:44:20] sftcd leaves the room
[20:44:21] <EKR > What the fuck is this shit?
[20:44:29] <Cullen Jennings> sort of agree with hta as long as poeple don't take the idea that no one stood up means we agree with it
[20:44:29] <Michael Richardson> Andrew McGregor: what if we dropped all packets in a flow label... that would hit an entire frame..? but not for tcp, I think.
[20:44:55] sftcd joins the room
[20:44:59] <EKR > chairs: why is this on the agenda?
[20:45:25] <michael.welzl> it gives context. one major goal is to create a mechanism that minimizes delay. this is about delay.
[20:45:55] <Ted.h> Should be on: http://www.ietf.org/proceedings/84/slides/slides-84-rmcat-3.pdf soon
[20:45:57] <EKR > could we also have a presentation about whether neutrinos are breaking the speed of light?
[20:45:57] <Michael Richardson> ekr: the point is that the network has problems that are not just "solved" by running TCP.
[20:46:24] <EKR > michael: well, basically nobody on the RAI side thinks you would just run TCP
[20:46:40] <michael.welzl> ekr: that is scheduled next :-)
[20:46:44] <Ted.h> Slide 2
[20:47:33] <Michael Richardson> EKR: yeah. but TCP's ability to "get out of the way" is constrained by bufferbloat.
[20:47:40] <Cullen Jennings> love this guy
[20:48:05] <michael.welzl> matt is giving context now.
[20:48:15] <EKR > I suppose. That could have been presented in like 1/4 the time
[20:48:17] <EKR > http://i0.kym-cdn.com/entries/icons/original/000/000/021/relevant-to-my-interests.jpg
[20:48:32] <Ted.h> Slide 4
[20:48:39] <JohnLeslie> EKR: not everybody can talk as fast as you ;^)
[20:49:19] <Cullen Jennings> We had to cut off EKR at his preso in rtcweb as he talked so slow, he went way over time budget
[20:49:19] <Michael Richardson> EKR does not suffer from bufferbloat induced delay.
[20:49:28] <EKR > Yeah, well, I was adding a lot of vlaue
[20:49:44] <Ted.h> slide 5
[20:51:18] <Ted.h> slide 6
[20:51:51] Zahed Sarker leaves the room
[20:51:59] <Ted.h> slide 7
[20:52:12] ena2413 joins the room
[20:53:01] <Ted.h> slide 8
[20:53:08] <rillian> so, what changed? the edges got faster faster than the backbone?
[20:53:35] <Ted.h> slide 9
[20:54:09] <EKR > ah
[20:55:09] <Ted.h> slide 10
[20:55:20] tsuichi leaves the room
[20:55:22] tsuichi joins the room
[20:56:50] <Ted.h> I heard Andrew say that as "It's not a good idea to be TCP-friendly, as it won't return the favor". It's pithy either way....
[20:56:51] martin.stiemerling leaves the room
[21:01:40] <Ingemar Johansson> MIC: Having a hard time to understand why not even an ideal AQM can solve things
[21:01:44] EKR leaves the room
[21:02:17] <rillian> amen
[21:02:34] <Ingemar Johansson> But no need to put it on the mic.., as Van says roughly the same
[21:02:47] <Cullen Jennings> that was tail drop on speaker q
[21:02:54] <Ingemar Johansson> :-)
[21:03:07] <Ted.h> Now Stefan Holmer's slides, Harald speaking
[21:03:20] EKR joins the room
[21:03:21] <Ted.h> Slide 2 (First comment slide)
[21:03:52] <Ted.h> Slide 3
[21:05:12] JohnLeslie leaves the room
[21:05:16] <Ted.h> Slide 4
[21:05:31] <EKR > kalman filter? Cullen Jennings to the microphone please!
[21:05:31] JohnLeslie joins the room
[21:05:54] <EKR > Did they do monte carlo markov chain simulations?
[21:06:05] <Ted.h> Slide 5
[21:06:06] <Cullen Jennings> A more accurate description is they are assuming a Gaussian distribution
[21:06:15] <EKR > I hear the bootstrap helps here.
[21:08:49] ena2413 leaves the room
[21:10:37] <Varun Singh> That is one reason RRTCC ignores packet loss upto 2%
[21:10:43] <Michael Richardson> Andrew: isn't this bit rate basically analogous to an advertised windows by a receiver?
[21:10:55] <Ted.h> Moving to Piers O'Hanlon's deck
[21:11:30] <Ted.h> Title slide
[21:11:35] ena2413 joins the room
[21:12:08] <Ted.h> slide 2
[21:12:17] <martin.thomson> bufferbloat
[21:13:00] <michael.welzl> chatbloat, slidebloat, talkbloat!
[21:13:16] <Michael Richardson> beerbloat.
[21:13:28] <michael.welzl> bloat-bloat!!!
[21:13:43] Michael Richardson sees a monty-python skit.
[21:14:08] <martin.thomson> it's waffer thin
[21:14:19] <Ingemar Johansson> You never expect the spanish bufferbloat
[21:14:44] <martin.thomson> http://www.youtube.com/watch?v=v29QfOyuZ3Y
[21:15:26] <Ingemar Johansson> OK, got it
[21:16:14] <hta> URL to materials:
[21:16:15] <Ted.h> slide 4
[21:16:15] <hta> https://datatracker.ietf.org/meeting/84/materials.html#session.group-rmcat
[21:17:15] Lars leaves the room
[21:17:57] <Ted.h> Now on "discussion slide"
[21:18:47] <Ted.h> Now on Proposed Charter
[21:19:16] <Cullen Jennings> can someone post the link to the charter ?
[21:19:34] <frodek> https://sites.google.com/a/alvestrand.com/rtp-congestion/bof-planning-page/wg-charter---input-to-vancouver
[21:19:39] <Ted.h> https://sites.google.com/a/alvestrand.com/rtp-congestion/bof-planning-page/wg-charter---input-to-vancouver
[21:19:47] <Ted.h> hmm over copy
[21:19:47] <michael.welzl> https://sites.google.com/a/alvestrand.com/rtp-congestion/bof-planning-page/wg-charter---input-to-vancouver [https://sites.google.com/a/alvestrand.com/rtp-congestion/bof-planning-page/wg-charter---input-to-vancouver]
[21:19:52] <Dan York> https://sites.google.com/a/alvestrand.com/rtp-congestion/bof-planning-page/wg-charter---input-to-vancouver
[21:20:26] <Dan York> Got the URL, Cullen? ;-)
[21:20:40] <EKR > http://rtp-congestion.alvestrand.com/bof-planning-page/wg-charter---input-to-vancouver
[21:21:02] <Cullen Jennings> by network has stalled loading them all at the same time
[21:23:39] EKR leaves the room
[21:24:07] <Jonathan Lennox> Milestones for when you want to have an I-D?
[21:24:18] <Ted.h> chartering plus 1
[21:24:27] <Ted.h> Now discussing charter
[21:25:20] rillian leaves the room
[21:25:46] ena2413 leaves the room
[21:25:51] hta leaves the room
[21:26:34] ena2413 joins the room
[21:27:20] frodek leaves the room
[21:28:57] <Ted.h> This extended metaphor is a little odd.
[21:29:34] <martin.thomson> I'm waiting for the first reference to the tragedy of the commons
[21:30:14] <Ted.h> Too late; it was on one of the first slides
[21:30:19] <Ingemar Johansson> MIC: Have a question as regards to feedback of congestion signals, at least one proposal implies receiver based CC, making it unnecessary to feedback actual congestion. However a ConEx enabled sender needs feedback of actual congestion signals to function properly. Is this concern something for this charter ?
[21:31:14] <Ted.h> okay i am in line for you
[21:31:50] <martin.thomson> all packets are created equal, but some are more equal than others
[21:32:03] <Dan York> Note that there are 5 people in front of Ted :-)
[21:32:26] EKR joins the room
[21:33:19] wolfgang.beck01 joins the room
[21:33:34] Wolfgang Beck leaves the room
[21:33:42] mikeb joins the room
[21:35:27] <JohnLeslie> MIC: why do we need Experimental to deploy something in browsers?
[21:35:42] <Ted.h> okay, caught
[21:37:09] ena2413 leaves the room
[21:38:18] ena2413 joins the room
[21:38:21] sftcd leaves the room
[21:41:18] <JohnLeslie> Van++
[21:41:27] <Dan York> Ted at mic
[21:41:41] tterribe joins the room
[21:42:10] Ted.h leaves the room
[21:42:14] <Ingemar Johansson> OK, thanks
[21:42:20] frodek joins the room
[21:42:27] ena2413 leaves the room
[21:42:28] EKR leaves the room
[21:42:32] <Dan York> Ted has relayed both comments
[21:43:12] <Ingemar Johansson> Drama!!!
[21:43:23] <Ingemar Johansson> :-)
[21:43:42] Ted.h joins the room
[21:47:17] <Varun Singh> +1 to lars about evaluation.
[21:47:37] Dan York leaves the room
[21:50:21] michael.welzl leaves the room
[21:50:33] <martin.thomson> the IETF produces first-class APIs
[21:50:46] <Cullen Jennings> who is that speaking right now ?
[21:54:39] Ted.h leaves the room
[21:57:06] tomkrist leaves the room
[21:58:18] <cabo> Mirja Kuehlewind
[21:58:38] <Varun Singh> +1
[21:58:44] <Varun Singh> agres
[21:59:46] <Varun Singh> support
[22:00:27] <Ingemar Johansson> You are really having fun..
[22:00:34] martin.thomson leaves the room
[22:00:51] mreavy leaves the room
[22:00:51] Jonathan Lennox leaves the room
[22:00:52] xiaoqing.zhu leaves the room
[22:01:07] Ingemar Johansson leaves the room
[22:01:08] wolfgang.beck01 leaves the room
[22:01:10] mramalho@cisco.com leaves the room
[22:01:13] Michael Tüxen leaves the room
[22:01:21] tsuichi leaves the room
[22:01:22] Magnus leaves the room: I'm happy Miranda IM user. Get it at http://miranda-im.org/.
[22:05:29] Varun Singh leaves the room
[22:07:02] sftcd joins the room
[22:07:59] mikeb leaves the room
[22:07:59] sftcd leaves the room
[22:09:21] cabo leaves the room
[22:12:22] tterribe leaves the room
[22:12:51] tsuichi joins the room
[22:13:27] EKR joins the room
[22:13:52] Varun Singh joins the room
[22:14:06] tterribe joins the room
[22:15:01] Cullen Jennings leaves the room
[22:15:19] tomkrist joins the room
[22:15:37] mreavy joins the room
[22:15:50] Jonathan Lennox joins the room
[22:16:09] Jonathan Lennox leaves the room
[22:17:22] Matt Zekauskas leaves the room
[22:18:12] wolfgang.beck01 joins the room
[22:18:31] hta joins the room
[22:19:58] wolfgang.beck01 leaves the room
[22:30:28] EKR leaves the room
[22:43:17] EKR joins the room
[22:49:35] EKR leaves the room
[22:49:38] EKR joins the room
[22:51:52] Michael Richardson leaves the room
[23:05:43] cabo joins the room
[23:06:19] Matt Zekauskas joins the room
[23:14:32] tomkrist leaves the room
[23:16:14] EKR leaves the room
[23:17:23] cabo leaves the room
[23:17:26] hta leaves the room
[23:17:56] EKR joins the room
[23:24:50] EKR leaves the room
[23:26:19] cabo joins the room
[23:26:53] cabo leaves the room
[23:39:46] Lars joins the room
[23:40:26] Matt Zekauskas leaves the room
Powered by ejabberd Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!