[14:01:17] gorryf joins the room
[14:03:30] spencerdawkins joins the room
[14:04:18] <gorryf> Place your contact details here, in Etherpad:
[14:04:20] Kazunori Fujiwara joins the room
[14:04:28] <gorryf> https://etherpad.ietf.org:9009/p/notes-ietf-interim-2020-tsvwg-02?useMonospaceFont=true
[14:05:24] <spencerdawkins> Could someone post the webex? I seem to have lost it ...
[14:06:23] <Magnus Westerlund> https://ietf.webex.com/ietf/j.php?MTID=mfd465befde515396f184f5706e441d5b
[14:07:20] martin.duke joins the room
[14:07:22] ekr@jabber.org joins the room
[14:13:44] <spencerdawkins> (Thank you, Magnus!)
[14:14:54] <spencerdawkins> RLC coding is now RFCs. My work on this planet is done.
[14:19:03] <spencerdawkins> If WGLC was a defined part of the process, that would be a less painful conversation ...
[14:27:24] Jake joins the room
[14:31:39] <Jake> to speak, put +q in the main webex chat.  just wanted that captured so we can find it later...
[14:35:26] Mirja joins the room
[14:35:55] csperkins joins the room
[14:39:07] <Magnus Westerlund> Doesn't TCP have the concept of congestion events. So why would two CE marks arriving back to back cuase multiple congestion events? Or is the assumption here that there a randomly distributed and thus cause multiple congestion events?
[14:40:20] <spencerdawkins> Magnus, should that be a +q?
[14:40:20] slblake joins the room
[14:42:15] <Jake> does that plug into a tool or something?
[14:45:25] <Magnus Westerlund> The q+ or +q is only human interpreted. Thus, it doesn't matter.
[14:50:26] <ekr@jabber.org> I'm not sure why we are using +q. It seems like an unnecessary deviation from W3C convention
[14:51:05] <Jake> what's w3c convention?
[14:51:10] <ekr@jabber.org> q+/q-
[14:54:31] <Magnus Westerlund> I think it has to do simply with who wrote the IETF 107 guide who might not be aware of the details of the W3C usage but heard about it.
[14:54:43] <spencerdawkins> @ekr, I think that is just a matter of unfamiliarity with using WebEx for queue management. I noticed during the "official" meeting week that I had used both forms unintentionally - but, of course, I've probably never typed q+ or +q in a meaningful way previously.
Do I remember that W3C uses a bot for queue management using q+/q-? If we were doing that, a lot of things would get simpler.
[14:56:57] chi.jiun.su joins the room
[15:06:20] <Magnus Westerlund> Looking forward to get the SCTP NAT document done. I was TSV AD when this document was adopted in Behave WG.
[15:11:03] ekr@jabber.org leaves the room
[15:15:38] <spencerdawkins> Magnus - maybe the SCTP NAT draft will be done before Last Call on the QUIC NAT document starts :p
[15:20:39] <Magnus Westerlund> @spencer hahaha.
[15:30:10] Mirja leaves the room: Disconnected: Replaced by new connection
[15:30:10] Mirja joins the room
[15:36:33] bob.briscoe joins the room
[15:49:30] <bob.briscoe> @Magnus, only just joined jabber. In response to your earlier question "Doesn't TCP have the concept of congestion events," when packets are marked (some of which happen to be fragments), CE marks don't arrive back to back. They don't have to be randomly distributed (as they would be with RED, PIE, etc). They can be deterministically spaced (e.g. CoDel, COBALT). But in all cases, if flow B has twice as many packets running through the AQM as flow A, twice as many packets will be marked.
[15:49:52] <martin.duke> Would it be possible to map 3GPP QoS behaviors to existing DSCPs that are similar?
[15:53:59] <Magnus Westerlund> @bob: Okay, so there appear to be several aspects here. One is transport protocol behaviors when receiving CE marks, and how that results in congestion response. The second part I assume is that how one count CE marks.
[15:54:11] <Magnus Westerlund> @bob does byte vs packet counting here make a significant difference?
[15:59:51] <spencerdawkins> The other awesome part of diffserv/qci mapping is that the final answer may be "you guys should really be using diffserv" ... and I could live with that if it was written down.
[16:01:12] Kazunori Fujiwara leaves the room
[16:04:01] bob.briscoe leaves the room
[16:04:07] Magnus Westerlund leaves the room
[16:04:12] martin.duke leaves the room
[16:04:41] slblake leaves the room
[16:10:37] spencerdawkins leaves the room
[16:37:00] gorryf leaves the room
[16:46:55] gorryf joins the room
[16:56:54] Jake leaves the room
[17:00:45] gorryf leaves the room
[17:24:09] Mirja leaves the room: Disconnected: closed
[17:25:00] Mirja joins the room
[17:30:21] Mirja leaves the room
[17:37:02] gorryf joins the room
[17:46:34] gorryf leaves the room
[17:56:33] gorryf joins the room
[18:37:14] chi.jiun.su leaves the room
[19:19:10] gorryf leaves the room
[19:19:59] gorryf joins the room
[19:24:29] gorryf leaves the room
[22:51:19] csperkins joins the room
[22:51:36] csperkins leaves the room