IETF
ippm
ippm@jabber.ietf.org
Friday, July 29, 2022< ^ >
Joseph has set the subject to: IPPM Working Group
Room Configuration
Room Occupants

GMT+0
[16:30:41] <zulipbot> (Ozgur Bulbul) Yes we can!
[16:30:54] <zulipbot> (Fabio Bulgarella) Yes
[16:31:23] <zulipbot> (Mauro Cociglio) yes
[16:31:23] <zulipbot> (Eduard V) Vienna on the slide
[16:34:20] <zulipbot> (Tommy Pauly) Fixed!
[16:59:59] <zulipbot> (Rich Salz) Lies, more lies, and browser user-agent strings.
[17:00:42] <zulipbot> (Tommy Pauly) Are you in the right room, Rich?
[17:01:08] <zulipbot> (Rich Salz) meetecho thinks so.  maprg?
[17:01:48] <zulipbot> (Tommy Pauly) IPPM
[17:01:48] <zulipbot> (Rich Salz) hm, page wrong.  oh well.  cya
[17:15:13] <zulipbot> (Will Hawkins) We (at goresponsiveness) *also* have the same feedback and problem
[17:16:01] <zulipbot> (Will Hawkins) The presence of encryption in the protocol means that it is more difficult to measure on low-power (ie, CPU-limited) devices. We think that this problem is significant.
[17:22:43] <zulipbot> (Eduard V) looks like we lost the room
[17:23:09] <zulipbot> (Christoph Paasch) We can't see the room. - but we can hear
[17:23:10] <zulipbot> (Eduard V) I hear Al Morton
[17:24:14] <zulipbot> (Will Hawkins) Too bad we aren't in a conference with the world's foremost authorities on networks
[17:24:40] <zulipbot> (Tommy Pauly) We're back
[17:51:21] <zulipbot> (J Ignacio Alvarez-Hamelin) Just a question, it seems more a transport measurement, why it is on IPPM?
[17:51:52] <zulipbot> (Tommy Pauly) It's measuring network capacity, using higher level transports
[17:52:22] <zulipbot> (J Ignacio Alvarez-Hamelin) o.k., thanks!
[17:52:22] <zulipbot> (Tommy Pauly) Yup!
[17:55:46] <zulipbot> (J Ignacio Alvarez-Hamelin) Why not measure the delay to detect the goodput situation?
[17:56:04] <zulipbot> (J Ignacio Alvarez-Hamelin) (sorry, lot of questions :)
[17:56:17] <zulipbot> (Tommy Pauly) Do you mean to just go until the responsiveness stops changing?
[17:56:30] <zulipbot> (Tommy Pauly) I think it's because at the start it doesn't change until we hit full good put capacity
[17:56:45] <zulipbot> (J Ignacio Alvarez-Hamelin) More o less, do you know BBR, ir defines when you are in one or other situation.
[17:59:16] <zulipbot> (Marco Tollini) fast.com by Netflix has a "loaded latency" as well. It would be nice to understand how they implemented it as well
[18:00:48] <zulipbot> (Will Hawkins) Great question, Marco!
[18:18:33] <zulipbot> (J Ignacio Alvarez-Hamelin) A comment on the previous one: I think that Netflix are interesting their  needs, I mean the bandwidth of his content (based on their definitions) and not the general question.