Friday, July 20, 2018< ^ >
synp has set the subject to: SecEvent IETF 101 -
Room Configuration
Room Occupants

[13:14:12] 9rvRpuFY joins the room
[13:18:07] meetecho joins the room
[13:31:35] Bjorn Hjelm joins the room
[13:31:53] William Denniss joins the room
[13:34:14] joins the room
[13:34:20] has set the subject to: SecEvent IETF 101 -
[13:34:41] Phil Hunt joins the room
[13:35:32] John Bradley joins the room
[13:39:20] <William Denniss> The speaker's microphone doesn't seem to be working
[13:39:33] <William Denniss> is it on?
[13:39:41] <William Denniss> Thanks!
[13:47:38] Stephanie Huguenin joins the room
[13:48:38] <Phil Hunt> MIC: Correct the JTI is being used for filtering
[13:49:07] <> Phil, are you happy with Tony's statement?
[13:49:09] <Phil Hunt> MIC: JTI was being used to acknowledge the SETs received since SETs are identified by JTI.
[13:49:23] <Phil Hunt> Yes I am happy
[13:51:15] <Phil Hunt> MIC: Disagree.
[13:51:57] <Phil Hunt> MIC: This fundamentally changes the protocol
[13:52:51] <Phil Hunt> It is not double JTI
[13:53:31] <Phil Hunt> MIC:  This item was in flux and needs to be resolved
[13:54:40] <> Phil: "this" is double-jti, yes?
[13:54:58] <Phil Hunt> This may not work
[13:55:47] <Phil Hunt> I will do by text..  this was broken in oauth
[13:56:12] <Phil Hunt> MIC: These two parameters are flags for different things.
[13:56:27] <Phil Hunt> MIC: return immediately is used for http long polling
[13:56:33] <meetecho> Phil Hunt: you may want to try our self test for next time, to see why audio doesn't work:
[13:56:38] <Phil Hunt> MIC: Agreed this text was still in flux
[13:57:59] <Phil Hunt> MIC:  THis is clear.  THis is about a transmitter that has not yet received an ack for an event identified by a JTI
[13:58:38] <Phil Hunt> Annabelle is correct
[13:58:52] <> yeah, I wa about to say that annabelle just said basically that
[13:59:30] <Phil Hunt> MIC:  It is a multi-step process. The following POST acks the received events and may await the next group of SETs
[14:02:24] <Phil Hunt> No audio coming
[14:02:45] William Denniss leaves the room
[14:02:46] William Denniss joins the room
[14:02:49] <Phil Hunt> Will re-load my meetecho
[14:03:03] <William Denniss> I just lost audio as well.
[14:03:04] <> I think the queue line's mic is at a lower gain than others, but
probably not so low so as to not give you anything at all
[14:03:08] <meetecho> Looking into this
[14:03:14] <meetecho> (looks like the Streamer died)
[14:03:21] <> And Mike has been talking some too, so if you can't hear him...
[14:03:35] <Phil Hunt> not for the last few minutes
[14:03:48] <William Denniss> audio was great, then it cut completely I don't think it's the mic
[14:03:49] <> Annabelle says that if we say this should be done programaticaly, we
should specify how to do that
[14:03:50] Phil Hunt leaves the room
[14:03:53] <meetecho> Phil Hunt: it's not you, audio stream has died
[14:04:01] Phil Hunt joins the room
[14:04:10] <> Mike is effectively reading this for the first time and it's wildly
[14:04:15] <> general agreement
[14:04:22] <> let me know when the audio comes back
[14:04:28] <William Denniss> Thank you kaduk
[14:04:29] <> clerical: need to add a registry for the values
[14:04:32] <meetecho> Sure, we're working on this as I type
[14:04:49] <> you have video, though?
[14:04:51] <William Denniss> yes
[14:04:53] <Phil Hunt> :-)  Still dead after I reloaded.  Must be at the transmitter side
[14:04:58] <Phil Hunt> Have video
[14:05:16] <> Okay, I will not transcribe the slides :)
[14:05:39] <meetecho> Phil Hunt: yes, as I said the streamer died
[14:05:50] <meetecho> Should be back any second now though
[14:05:57] <William Denniss> it's back
[14:05:59] <Phil Hunt> Got it thanks!
[14:06:06] <> Thanks meetecho!
[14:06:09] <meetecho> (y)
[14:06:13] <William Denniss> Thank you meetecho for the fix, and Kaduk for the transcribing!
[14:06:58] <Phil Hunt> Ben - i am concerned now that the specs are split that they won't stay in sync.  There is a push to move push ahead without waiting. This is why I resigned as editor and strongly objected to the consensus.  FYI
[14:07:47] <> Thanks, Phil
[14:10:42] <Phil Hunt> MIC: Anything that requires registry (e.g. errors) is a problem to duplicate
[14:10:53] <> yeah
[14:11:50] <Phil Hunt> MIC: PS. From annabelles presetation, the use of http status is to follow normal htpp. Use of error codes is due to message processing like is the JWT valid.
[14:12:53] <Phil Hunt> correct
[14:12:57] <Phil Hunt> thanks!
[14:15:01] <Phil Hunt> MIC: Or since only a couple of implementations exist consider a new proposal (SSTP) which is similar to both and unifies the method
[14:15:25] <Phil Hunt> MIC: Only  one has implemented a push transmitter
[14:17:02] <Phil Hunt> MIC: I disagree with annabelle. The whole point of forming SECEVENTs was to develop a universal delivery method
[14:20:48] <Phil Hunt> MIC: That feels like NIMBYism on requirements.
[14:21:00] <William Denniss> Dick makes a good point about OAuth, a lot of people get scared even though they may just need the implicit flow which is a single HTTP request/response.
[14:21:59] <Phil Hunt> MIC: The group could simply solicit new proposals. It does not have to take sstp
[14:22:15] <Phil Hunt> single
[14:32:38] <> The "sub" claim, or the sub-claim? ;)
[14:32:53] <Phil Hunt> hahah... "sub" claim
[14:33:04] <Phil Hunt> It is defined in the JWT spec
[14:33:48] <Phil Hunt> Problem for SET was "sub" was often bound to an "iss", but in SET, the "iss" may not be the issuer of the "sub"  :-)
[14:34:07] <> Yeah, I remember the distinction, just when you're saying it out loud
you can hear it either way
[14:34:16] <Phil Hunt> I see
[14:36:23] <> Remind me what OP is?
[14:36:38] <Phil Hunt> An "OpenID" Provider
[14:36:45] <> Okay, thanks.  
[14:36:48] Bjorn Hjelm leaves the room
[14:36:58] <> That would have been my guess, but this isn't the OpenID group :)
[14:38:01] Bjorn Hjelm joins the room
[14:38:20] <Phil Hunt> MIC: It can be space constrained as SETs are JWTs and may need to be passed in constricted places
[14:39:45] <Phil Hunt> MIC: I am also hearing that some SET scenarios are being used for attestation. Using URIs really expands the size and costs of retention.
[14:47:18] <> Sorry it took so long to get up there!
[14:47:29] Stephanie Huguenin leaves the room
[14:47:54] <Phil Hunt> Not a problem.  Sorry meetecho isn't liking my microphone.
[14:49:08] <> Call the Internet Police!
[14:49:34] <> "nobody cares about LoA" from the back of the room
[14:51:35] <Phil Hunt> Agreed
[14:53:07] <Phil Hunt> MIC: Regarding OAUth token revoke, I found this very confusing when I read it.
[14:53:09] Bjorn Hjelm leaves the room
[14:53:59] <Phil Hunt> MIC: Also, I would prefer the OAuth WG define token revoke event.
[14:56:11] Bjorn Hjelm joins the room
[14:56:29] William Denniss leaves the room
[14:58:08] <Phil Hunt> MIC: I have several times
[15:01:50] <Phil Hunt> ( need to say my last comment...we can move on)
[15:04:08] <> I was going to do it out-of-band-ish, but thanks for confirming
[15:04:29] John Bradley leaves the room
[15:08:18] <Phil Hunt> Thanks for being my relay ben!
[15:08:33] Phil Hunt leaves the room
[15:09:04] Bjorn Hjelm leaves the room
[15:09:11] William Denniss joins the room
[15:10:28] <> You're welcome!
[15:10:52] <William Denniss> Bye all, thanks Ben for transcribing when the audio was cut :)
[15:11:17] meetecho leaves the room
[15:11:23] William Denniss leaves the room
[15:56:17] leaves the room
Powered by ejabberd - robust, scalable and extensible XMPP server Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!