IETF
QUIC
quic@jabber.ietf.org
Wednesday, November 7, 2018< ^ >
Christian has set the subject to: Siren Retry
Room Configuration
Room Occupants

GMT+0
[00:33:36] ekr joins the room
[01:47:17] Thomas Peterson joins the room
[01:54:28] Meetecho joins the room
[01:54:52] Eric Kinnear joins the room
[01:55:04] Simon Pietro Romano joins the room
[01:55:08] Dmitri Tikhonov joins the room
[01:56:03] Alan Frindell joins the room
[01:56:05] Bernard Aboba joins the room
[01:56:34] Martin Thomson joins the room
[01:57:09] ekr leaves the room
[01:58:12] Dan York joins the room
[01:58:16] Philipp Tiesel joins the room
[01:58:18] Lars Eggert joins the room
[01:59:14] Dragana Damjanovic joins the room
[02:01:00] chi.jiun.su joins the room
[02:02:23] Kurt Andersen joins the room
[02:03:02] Erik Nygren joins the room
[02:03:06] Victor Vasiliev joins the room
[02:03:32] Hirochika Asai joins the room
[02:04:31] ted.h joins the room
[02:04:34] Magnus Westerlund joins the room
[02:04:46] Mat Ford joins the room
[02:04:46] <ted.h> If you would like your point reflected to room, please put MIC: in front.
[02:05:11] ekr joins the room
[02:06:02] nygren joins the room
[02:06:33] vojislav vucetic joins the room
[02:07:42] DavidSchinazi joins the room
[02:07:43] Bo Burman joins the room
[02:08:02] terrebyte joins the room
[02:08:41] <ted.h> And first, a digression about beer….sorry, I meant an argument about beer.
[02:08:48] blassey joins the room
[02:09:40] frodek joins the room
[02:10:10] dkg joins the room
[02:10:18] <ted.h> For folks entering now, if you would like your point reflected to room, please put MIC: in front.
[02:11:13] DavidSchinazi leaves the room: Stream closed by us: Replaced by new connection (conflict)
[02:11:24] DavidSchinazi joins the room
[02:11:39] nat joins the room
[02:21:38] Jonathan Lennox joins the room
[02:21:45] Kaoru Maeda joins the room
[02:23:42] <Martin Thomson> disabling ACK delay might help to confirm that hypothesis
[02:24:58] Lars Eggert leaves the room: Disconnected: Replaced by new connection
[02:25:00] Lars Eggert joins the room
[02:27:36] Bernard Aboba leaves the room
[02:27:36] Kaoru Maeda leaves the room
[02:28:35] Bernard Aboba joins the room
[02:29:48] <ted.h> Line currently at 2.
[02:31:00] Nicholas Banks joins the room
[02:31:48] Mat Ford leaves the room
[02:33:11] Suzanne Woolf joins the room
[02:34:37] Mat Ford joins the room
[02:34:50] nat leaves the room
[02:35:00] ekr leaves the room
[02:36:47] sftcd joins the room
[02:37:06] ekr joins the room
[02:37:36] Samuel Jero joins the room
[02:37:48] <ted.h> Can we bring back the slide with the beer?  Asking for a friend.
[02:39:46] resnick joins the room
[02:40:50] Mat Ford leaves the room
[02:41:02] <ted.h> Strong hum for limit to single bit variant
[02:41:02] g.e.montenegro joins the room
[02:41:58] <resnick> Data point from the back of the room: There was a tiny little hum for other than single bit, but it was non-committal and I think the chairs confirmation of the sense of the room is still correct.
[02:42:49] <sftcd> so it wasn't you then Pete? :-)
[02:43:16] <resnick> I sit here, eating my popcorn, humlessly.
[02:44:09] Dmitri Tikhonov hums for "fit for purpose"
[02:44:13] <resnick> Full cricketness.
[02:44:32] <nygren> hums for "fit for purpose"
[02:45:28] <ekr> It’s possible not to spin if you are a boson, but not if you are a fermion
[02:47:26] <Jonathan Lennox> If this is your first time at spin club, you must spin
[02:48:22] <resnick> Dead Or Alive - You Spin Me Round (Like a Record) (Official Video ... <https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact=8&ved=2ahUKEwjc-Z-wosHeAhWHpo8KHT4rDccQyCkwAHoECAQQBQ&url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DPGNiXGX2nLU&usg=AOvVaw0sIPV6jtbTZ1_xGo700gwH>
[02:51:25] <ekr> I actually think Christian’s point is pretty important
[02:51:42] Philipp Tiesel humms in favor of "opt out must be possible without consequences"
[02:51:55] <ekr> Or rather Marten’s
[02:52:00] Mat Ford joins the room
[02:55:08] Mat Ford leaves the room
[02:55:20] <ekr> Given the proposed designs, I am not sure how to deal with Marten’s concern about differential treatment by the network
[02:56:29] Lars Eggert leaves the room: Disconnected: closed
[02:56:32] Lars Eggert joins the room
[02:56:56] Philipp Tiesel leaves the room
[02:58:23] Vidhi Goel joins the room
[02:58:40] Philipp Tiesel joins the room
[02:59:30] Mat Ford joins the room
[03:00:33] resnick leaves the room
[03:00:35] resnick joins the room
[03:00:50] <resnick> I think we’re being asked whether we’ve got enough to go forward and work on this. I don’t think we’re being asked whether it’s possible we’ll find a showstopper at some point in the future. Correct?
[03:01:05] <ekr> I can’t quite tell.
[03:01:10] <ekr> That’s what I was trying toa sk
[03:01:48] Mat Ford leaves the room
[03:02:19] Mat Ford joins the room
[03:03:04] Vidhi Goel leaves the room
[03:03:10] Vidhi Goel joins the room
[03:03:30] <resnick> @ekr: If it is the first (i.e., “We are good to start working on this”), you’re OK with going forward?
[03:03:36] <ted.h> I tend to agree with Marcus; since the server can also opt out, it would be really, really weird for an operator to penalize folks in the presence of opt out.
[03:04:01] <ekr> @resnick: not necessarily
[03:04:07] <resnick> Interesting.
[03:04:16] <ted.h> I really haven't seen any compelling reason for them to compel.
[03:04:36] <ekr> @hardie: note that this is a non-technical argument, but rather an incentive argument
[03:04:52] Mat Ford leaves the room
[03:05:16] <ekr> Consider the case where everyone is spinning. It wouldn’t be unreasonable for the operator to degrade behavior in order to disincentivize that behavior
[03:05:47] <ted.h> @ekr Sorry.  It degrades behavior to stop spinning?  Why?
[03:05:59] <ekr> Because it wants an equilibrium where everyone is spinning
[03:06:12] <ekr> And wants to disincentive people from implementing non-spin clients
[03:06:52] <ted.h> @ekr Ah, so I misinterpreted you.  I thought you meant it was degrading spinning behavior, in order to stop it.
[03:06:56] <sftcd> as a nit: the term "opt out" implies defaulting to spinning which I don't think is quite whats meant
[03:07:18] <ekr> @sftcd: that was already decided! STFU!!!!!
[03:08:55] Mat Ford joins the room
[03:08:55] Lars Eggert leaves the room: Disconnected: closed
[03:08:56] <sftcd> I'm ok with the 1st use of it, the one just added to the slide might be an issue though
[03:09:49] Lars Eggert joins the room
[03:12:19] Mat Ford leaves the room
[03:12:53] <Martin Thomson> wow, ALPN would be a spectacularly inappropriate mechanism to use
[03:13:25] <sftcd> +1
[03:13:56] Mat Ford joins the room
[03:14:04] jhoyla joins the room
[03:14:35] <Philipp Tiesel> +1 (HQ+SPIN+FOO+UNICORN)
[03:15:37] <ted.h> Not an invariant, so other apps could make other decisions…
[03:15:47] Mat Ford leaves the room
[03:16:09] <Martin Thomson> THEN MOVE ON PLEASE
[03:16:10] <jhoyla> Is it not possible to emulate any arbitrary RTT? If you can, then it’s presumably impossible to discriminate between true spinning and emulated spinning.
[03:16:28] <ekr> I’m nappy to move on to the next question
[03:16:28] <ted.h> @jhoyla  You can emulate a longer one, not a shorter one.
[03:16:42] <ted.h> @jhoyla  That's why the VPN issue is there—you can
[03:16:45] <ekr> Mark asked people to get up, hence I did
[03:16:56] <ted.h> not emulate the RTT of being not behind a VPN.
[03:17:13] Mat Ford joins the room
[03:17:20] <jhoyla> @hardie Interesting, thanks.
[03:17:30] Praveen Balasubramanian joins the room
[03:17:39] <ekr> @hardie: are you sure you can’t emulate a shorter one
[03:18:34] <ekr> Suppose that I have perfect knowledge of the traffic patterns (e.g., we transmit a packet every 1 ms, but the latency is 5 ms)
[03:18:46] <ekr> So ordinarily, this would cause a flip every 5s
[03:19:19] <ekr> Just in case people didn’t notice, when I told Stephen above to STFU, that was intended to be joking. Should have included a smiley
[03:19:44] <sftcd> I always interpret being told that with a smiley, so it's fine:-)
[03:19:46] <Praveen Balasubramanian> Confirming again: Microsoft  will deploy if included in spec
[03:20:00] <ted.h> @Praveen  Do you want that said to the MIC?
[03:20:19] <Praveen Balasubramanian> I think Gabriel said that  just now so covered thanks
[03:20:37] <Dmitri Tikhonov> MIC: LiteSpeed will implement in our web server and proxy as well as our client library.
[03:21:22] <nygren> (Context: I believe 3GPPP is also looking at using QUIC for infrastructure signaling.)
[03:22:26] <jhoyla> @ekr @hardie If I just flip every 3 seconds, ignoring the true RTT, what does that look like on the network?
[03:23:29] <ekr> @jhoyla, I think you would have to be a little more sophisticated, perhaps
[03:23:34] <ekr> but yes, I think you could do that
[03:23:37] jhoyla leaves the room: Stream closed by us: Replaced by new connection (conflict)
[03:23:40] jhoyla joins the room
[03:24:01] Suzanne Woolf leaves the room
[03:24:25] resnick leaves the room
[03:24:27] <ted.h> @jhoyla  You can't do it quite that naively, because you have to take the echo into account.  To make this a true emulation, you need what ekr implied:  knowledge of the RTT for the device location you want to emulate.
[03:26:58] <jhoyla> @hardie and if I just wanted to set an RTT that was arbitrary, i.e. not related in any way to my true RTT, rather than a specific one, would that require the same information?
[03:27:39] <ted.h> @jhoyla I think that would make you stand out much more than simply failing to spin.  Personal opinion, of course.
[03:28:55] Mat Ford leaves the room
[03:28:56] kaduk@jabber.org/barnowl joins the room
[03:32:59] Mat Ford joins the room
[03:33:05] <Praveen Balasubramanian> hum for inclusion
[03:33:06] <Nicholas Banks> hum for including spin bit
[03:33:08] Dmitri Tikhonov hums for inclusion
[03:33:14] <Erik Nygren> hum for inclusion
[03:33:21] Philipp Tiesel hums for inclusion
[03:34:12] <kaduk@jabber.org/barnowl> Interestingly, house left was strong "yes" and all the "no"s seemed to
be coming from farther away.
[03:34:26] <ted.h> Rough consensus to include the spin bit.
[03:37:47] Viktor Dukhovni joins the room
[03:39:29] <kaduk@jabber.org/barnowl> Wow, I need to remember to give that guidance on referring to previous
speakers' statements in all my WGs
[03:39:32] <sftcd> +1 for not negotiating this
[03:39:53] <Erik Nygren> humm for not negotiating this
[03:40:01] <Philipp Tiesel> humm for not negotiating this
[03:41:26] <ted.h> +1 to Christian.  Good process around that are key.
[03:43:27] jhoyla leaves the room: Stream closed by us: Replaced by new connection (conflict)
[03:44:53] DavidSchinazi leaves the room
[03:44:55] DavidSchinazi joins the room
[03:45:21] <ekr> i mean, we’ll probably just have to move the deadline again, but who cares
[03:45:51] <Praveen Balasubramanian> Any discussion on removing some features to accelerate the schedule for v1?
[03:46:01] <ekr> I don’t think that’s helpful
[03:46:07] <Praveen Balasubramanian> ok
[03:46:19] <ekr> Most of the stuff that’s dragging us down is not really features (except maybe migration)
[03:46:37] <Praveen Balasubramanian> yeah that's the one I think we can live without in v1
[03:48:08] jhoyla joins the room
[03:49:17] Mat Ford leaves the room
[03:50:23] ekr leaves the room
[03:50:27] Mat Ford joins the room
[03:50:41] Giovane C. M. Moura joins the room
[03:51:11] <Eric Kinnear> I don't know that we can, but I agree that the stuff dragging us down isn't really features (and echoing the comments from the previous session about churn and the other factors influencing this)
[03:53:19] jhoyla leaves the room
[03:53:31] ekr joins the room
[03:55:25] Mat Ford leaves the room
[03:55:59] Mat Ford joins the room
[03:57:06] Mat Ford leaves the room
[03:57:07] <ekr> I think this point about handshake completion is actually wrong
[03:57:31] <ekr> Because of the design of DTLS, the handshake can be *more* aggressive than with QUIC
[03:59:51] sftcd leaves the room
[04:00:10] ekr leaves the room
[04:00:21] ted.h leaves the room
[04:00:22] Jonathan Lennox leaves the room
[04:00:27] blassey leaves the room
[04:00:28] DavidSchinazi leaves the room: Stream reset by peer
[04:00:28] Kurt Andersen leaves the room
[04:00:34] Lars Eggert leaves the room: Disconnected: closed
[04:00:36] Eric Kinnear leaves the room: Stream reset by peer
[04:00:39] Martin Thomson leaves the room
[04:00:46] Nicholas Banks leaves the room
[04:00:51] nygren leaves the room
[04:00:53] Alan Frindell leaves the room
[04:00:58] Simon Pietro Romano leaves the room
[04:00:58] Bo Burman leaves the room
[04:00:58] Praveen Balasubramanian leaves the room
[04:01:02] Philipp Tiesel leaves the room
[04:01:03] Dragana Damjanovic leaves the room
[04:01:03] Giovane C. M. Moura leaves the room
[04:01:03] Viktor Dukhovni leaves the room
[04:01:03] Erik Nygren leaves the room
[04:01:03] Dmitri Tikhonov leaves the room
[04:01:03] Samuel Jero leaves the room
[04:01:03] Bernard Aboba leaves the room
[04:01:03] Vidhi Goel leaves the room
[04:01:03] vojislav vucetic leaves the room
[04:01:03] Hirochika Asai leaves the room
[04:01:03] Victor Vasiliev leaves the room
[04:01:22] Meetecho leaves the room
[04:02:42] Thomas Peterson leaves the room: Disconnected: closed
[04:09:13] Mat Ford joins the room
[04:09:50] frodek joins the room
[04:10:56] 1g1sFs63 leaves the room
[04:12:22] frodek leaves the room
[04:13:59] Kurt Andersen joins the room
[04:15:08] Dan York joins the room
[04:15:09] Dan York leaves the room
[04:16:09] Mat Ford leaves the room
[04:16:13] Thomas Peterson joins the room
[04:16:15] Thomas Peterson leaves the room
[04:16:31] frodek leaves the room
[04:16:35] Dan York leaves the room
[04:18:58] g.e.montenegro leaves the room
[04:20:16] Mat Ford joins the room
[04:20:26] Martin Thomson joins the room
[04:21:01] Jonathan Lennox joins the room
[04:21:19] Jonathan Lennox leaves the room
[04:21:26] ekr joins the room
[04:21:29] DavidSchinazi joins the room
[04:22:10] Mat Ford leaves the room
[04:22:14] DavidSchinazi leaves the room: Stream closed by us: Replaced by new connection (conflict)
[04:22:24] DavidSchinazi joins the room
[04:22:58] Eric Kinnear joins the room
[04:23:07] Magnus Westerlund leaves the room: Stream closed by us: Replaced by new connection (conflict)
[04:23:08] Magnus Westerlund joins the room
[04:23:21] Mat Ford joins the room
[04:24:55] blassey joins the room
[04:29:14] DavidSchinazi leaves the room
[04:29:17] g.e.montenegro joins the room
[04:29:49] Mat Ford leaves the room
[04:30:41] g.e.montenegro leaves the room
[04:35:49] Eric Kinnear leaves the room: Stream closed by us: Replaced by new connection (conflict)
[04:35:59] Eric Kinnear joins the room
[04:41:24] blassey leaves the room
[04:41:32] ekr leaves the room
[04:47:51] Eric Kinnear leaves the room: Stream reset by peer
[04:49:57] ekr joins the room
[04:50:59] ekr leaves the room
[04:53:18] ekr joins the room
[04:54:07] ekr leaves the room
[04:55:11] Eric Kinnear joins the room
[04:55:55] Mat Ford joins the room
[04:56:28] Mat Ford leaves the room
[04:57:49] ekr joins the room
[04:59:43] ekr leaves the room
[05:04:26] ekr joins the room
[05:08:58] ekr leaves the room
[05:09:19] Eric Kinnear leaves the room: Stream reset by peer
[05:09:37] Martin Thomson leaves the room
[05:12:22] Kurt Andersen leaves the room
[05:14:44] ekr joins the room
[05:17:15] ekr leaves the room
[05:19:47] Kurt Andersen joins the room
[05:26:33] Mat Ford joins the room
[05:31:32] chi.jiun.su leaves the room
[05:36:01] Magnus Westerlund leaves the room: Stream closed by us: Replaced by new connection (conflict)
[05:36:03] Magnus Westerlund joins the room
[05:36:42] Mat Ford leaves the room
[05:53:51] Magnus Westerlund leaves the room: Connection failed: connection timed out
[05:56:27] Martin Thomson joins the room
[06:00:26] dkg leaves the room
[06:18:02] kaduk@jabber.org/barnowl leaves the room
[06:21:32] Martin Thomson leaves the room
[06:48:54] ekr joins the room
[06:50:52] Martin Thomson joins the room
[07:03:46] Eric Kinnear joins the room
[07:07:52] ekr leaves the room
[07:11:16] ekr joins the room
[07:32:15] Brian joins the room
[07:32:44] Magnus Westerlund joins the room
[07:34:32] ekr leaves the room
[07:36:02] Martin Thomson leaves the room
[07:37:20] ekr joins the room
[08:06:11] ekr leaves the room
[08:06:53] jhoyla joins the room
[08:07:22] ekr joins the room
[08:09:31] Martin Thomson joins the room
[08:14:54] jhoyla leaves the room
[08:16:06] ekr leaves the room
[08:19:46] Martin Thomson leaves the room
[08:22:48] Eric Kinnear leaves the room: Stream reset by peer
[08:25:14] terrebyte leaves the room
[08:25:15] terrebyte joins the room
[08:25:46] Brian leaves the room: Disconnected: closed
[08:28:51] Eric Kinnear joins the room
[08:30:46] Brian joins the room
[08:54:09] jhoyla joins the room
[09:15:38] Magnus Westerlund leaves the room: Connection failed: connection timed out
[09:18:22] terrebyte leaves the room
[09:18:25] terrebyte joins the room
[09:19:55] jhoyla leaves the room
[09:31:18] ekr joins the room
[09:31:41] jhoyla joins the room
[09:32:24] jhoyla leaves the room: Stream closed by us: Replaced by new connection (conflict)
[09:32:25] jhoyla joins the room
[09:33:10] ekr leaves the room
[09:37:52] ekr joins the room
[10:01:29] Eric Kinnear leaves the room: Stream reset by peer
[10:12:21] jhoyla leaves the room: Connection failed: connection closed
[10:15:14] ekr leaves the room
[10:20:11] Magnus Westerlund joins the room
[10:25:10] Magnus Westerlund leaves the room
[10:27:23] Martin Thomson joins the room
[10:27:25] Martin Thomson leaves the room
[10:37:25] ekr joins the room
[10:42:49] Eric Kinnear joins the room
[10:47:49] jhoyla joins the room
[11:02:04] jhoyla leaves the room
[11:32:52] ekr leaves the room
[11:36:29] ekr joins the room
[12:10:54] ekr leaves the room
[12:11:38] ekr joins the room
[12:12:00] ekr leaves the room
[13:08:10] Eric Kinnear leaves the room: Stream reset by peer
[13:11:21] Eric Kinnear joins the room
[13:11:41] terrebyte leaves the room
[13:11:43] terrebyte joins the room
[13:13:26] Eric Kinnear leaves the room: Stream reset by peer
[13:14:09] Eric Kinnear joins the room
[13:15:42] ekr joins the room
[13:16:42] Eric Kinnear leaves the room: Stream reset by peer
[13:18:42] ekr leaves the room
[13:46:05] terrebyte leaves the room
[13:46:07] terrebyte joins the room
[13:56:24] terrebyte leaves the room
[14:07:21] terrebyte joins the room
[14:13:32] ekr joins the room
[14:16:29] terrebyte leaves the room
[14:16:33] terrebyte joins the room
[14:25:56] terrebyte leaves the room
[14:26:03] terrebyte joins the room
[15:21:47] Brian leaves the room: Disconnected: closed
[17:06:54] Kurt Andersen leaves the room
[17:13:27] Kurt Andersen joins the room
[17:14:54] ekr leaves the room
[19:14:26] Kurt Andersen leaves the room
[19:18:53] Kurt Andersen joins the room
[19:30:56] Kurt Andersen leaves the room
[20:06:11] Kurt Andersen joins the room
[20:18:26] Kurt Andersen leaves the room
[20:33:19] Brian joins the room
[20:33:29] Brian leaves the room: Disconnected: closed
[20:52:58] terrebyte leaves the room
[20:53:01] terrebyte joins the room
[22:06:50] Kurt Andersen joins the room
[22:18:55] Kurt Andersen leaves the room
[22:57:54] ekr joins the room
[23:24:27] Kurt Andersen joins the room
[23:36:56] Kurt Andersen leaves the room
Powered by ejabberd - robust, scalable and extensible XMPP server Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!