IETF
radext
radext@jabber.ietf.org
Friday, April 8, 2016< ^ >
hartmans has set the subject to: IETF 92 radext meeting
Room Configuration
Room Occupants

GMT+0
[14:59:26] jimsch joins the room
[15:12:03] Meetecho joins the room
[15:16:48] Lorenzo Miniero joins the room
[15:17:08] <Meetecho> audio not up yet in meetecho, fixing that
[15:18:56] Alan DeKok joins the room
[15:26:54] <Alan DeKok> for the data types doc, the bulk of the conversion was done via a script.
[15:27:19] <Alan DeKok> the simple data types should be OK.  the complex data types (e.g. string) require a bit more review
[15:27:49] <Alan DeKok> along with data types I picked because the original RFCs were wrong, incomplete, or contradictory
[15:29:30] <Alan DeKok> I can't get on Jabber, so I hope someone is monitoring the meetocho chat
[15:29:51] <jimsch> Alan:  It is the same thing
[15:30:02] <Alan DeKok> thanks
[15:31:47] <jimsch> meetecho:  our second screen is not projecting
[15:32:06] <Meetecho> ack, looking into that
[15:32:25] <Alan DeKok> Lionel: Yes, names are not managed at all by IANA
[15:33:00] <jimsch> let me know if you want something relayed
[15:33:06] <Alan DeKok> sure
[15:35:52] <Alan DeKok> Meetocho: If I'm a speaker, I don't have a "raise hand" button to join the queue.  If I'm not a speaker, it's there
[15:42:10] <jimsch> alan:  Personally I always perfer positive to negative statements esp for 2119 langauge
[15:42:22] <Alan DeKok> makes sense
[15:45:54] <jimsch> The CoA proxy MUST send the NAS an attribute in the packet, only when the NAS did ...
[15:46:47] <Alan DeKok> that's really it... the CoA proxy MAY send the NAS attributes which the NAS sent to a RADIUS server.
[15:46:56] <Alan DeKok> i.e. it can send them, but doesn't have to send all of them
[15:47:16] Ville Hallivuori joins the room
[15:47:19] <Alan DeKok> I'll try to word-smith something that's clear and makes sense
[15:47:40] <jimsch> Yeah I need to re look at the text.   It disappeared in the middle of my trying to re-word it
[15:48:34] <Alan DeKok> The CoA proxy MUST ensure that the only attribute(s) it sends to a NAS are attributes which the NAS sent to a RADIUS server
[15:48:52] <Alan DeKok> mostly... except for change of authorization, where you're telling it to do something new
[15:49:20] <Alan DeKok> it's probably better to just say "when a RADIUS proxy adds X outbound, the CoA proxy MUST remove X on the reverse path"
[15:49:32] <jimsch> /are/are in response to/ ????
[15:49:54] <jimsch> Yes that makes sense
[15:52:37] Lorenzo Miniero leaves the room
[15:59:18] Lorenzo Miniero joins the room
[16:22:40] Ville Hallivuori leaves the room
[16:26:37] jimsch leaves the room
[16:27:14] Lorenzo Miniero leaves the room
[16:27:49] Meetecho leaves the room
[16:27:51] Alan DeKok leaves the room