[06:58:52] burn joins the room [07:01:10] mlm.michael.miller joins the room [07:01:21] Otmar Lendl joins the room [07:03:35] good morning drinks! [07:04:46] good morning [07:05:11] anyone here who is not in the room? [07:05:37] * Agenda bashing [07:05:39] as0-d91k joins the room [07:05:50] I am not in the room (Dan Burnett) [07:06:11] thanks for your various RAI jabber scribing :) [07:06:39] I'll be your jabber scribe for today, please indicate if you want me to bring a Q to the mic [07:06:50] no objections to agenda [07:07:12] Gonzalo joins the room [07:07:19] document status [07:07:58] alex asks for reviewers of usecase-requirements [07:08:11] btw, agenda is here http://tools.ietf.org/wg/drinks/agenda [07:08:27] Jean-Francois M. joins the room [07:08:52] we're at http://tools.ietf.org/agenda/78/slides/drinks-4.ppt [07:09:09] milestones... [07:09:49] last call for reqs soon, if we get reviewers [07:11:10] rich will step down as chair. [07:11:23] any volunteers? [07:11:40] next agenda point: use cases [07:12:05] http://tools.ietf.org/agenda/78/slides/drinks-0.ppt [07:12:14] sumanth presenting [07:14:41] slide #5 [07:16:48] who is talking? [07:17:12] Mr Mule: policy vs. organizations. [07:17:23] thx [07:17:34] jean francois [07:17:41] david schwarz now. [07:17:50] tomkri joins the room [07:18:08] david: policy info is different from routing info [07:19:56] jean-francois, very simple policy. the grouping is enough [07:20:46] if there are multiple routes, selection is up to the data recipient, not the registry [07:21:14] jon peterson [07:22:15] sayed alid: [07:22:43] david answering [07:23:45] david: terminating side will also want to make decisions [07:24:57] jf: both originator and terminator are able to make choices [07:25:27] jf: terminating side can introduce multiple routes for a destination into the system [07:27:12] alex: there can be many routes in the system (hundreds), a given SSP might see a dozen of those [07:31:40] me on the mic [07:32:46] daryl: traffic engineering with this protocol should be possible [07:34:11] brian rosen: change the uses-case "delegated name server" to "refernce to some other data-source" [07:34:53] david: routing decision can by quite dynamic [07:34:58] jgunn joins the room [07:37:38] otmar : should policies by in the registry, or should policies lead to provisioning action towards the registry [07:38:00] jf: no need for a policy language [07:38:20] alex: perhaps transport policy, but not execute it [07:38:28] jon: agrees [07:38:40] slide . [07:38:43] 6. [07:43:17] otmar: do we need sip:*@domain as a wildcard PI [07:45:21] daryl: why not use 3263 for uri-based dialling [07:45:49] answers: because direct 3263 connectivity is the exception, not the rules [07:46:13] slide 7 [07:47:12] slide 8 [07:50:50] slide 9: direct PI-SED link perhaps used for customer-run endpoints [07:51:51] alex asks for reviewers AGAIN. [07:52:38] sohel [07:52:41] jon [07:52:54] we've got two. [07:53:38] next agenda item. [07:53:45] Ken presenting [07:54:09] http://tools.ietf.org/agenda/78/slides/drinks-1.pptx [07:54:30] Ken is slightly quiet on audio [07:55:35] is that better? [07:55:36] yes, audio is better now. thx [07:57:22] route group offer/accept concet [07:59:00] david: notification scheme? [07:59:24] ken: poll queue or push mechanism? [07:59:48] ken: epp uses polling, so perhaps better than push [08:02:52] ken: push is much more complex [08:03:20] jm: not in the first version of the protocol [08:06:25] carrier of record claim (slide 6) [08:09:38] Otmar asked about how the COR claim is adressed in the use case document [08:10:04] Ken explained that it is tied in with the transit vs. cor provider [08:10:50] slide: next steps [08:12:47] next agenda item: syed presenting [08:13:04] http://tools.ietf.org/agenda/78/slides/drinks-5.pptx [08:14:04] Otmar on slide 1: why email? [08:15:38] jm: perhaps return the email, not query for the email [08:17:13] suzukisn joins the room [08:17:26] suzukisn leaves the room [08:18:01] Jon: the charter was written with ENUM in mind: query with a TN, output many things out [08:18:28] so returning a URI is ok [08:18:36] => change "email" to URI [08:19:21] and Otmar clarified this is not intended to be a param for query but it is a param to be returned [08:20:14] david on the mike [08:21:15] usage scenario (slide 4) [08:21:58] slide 5 [08:24:15] slide 6 [08:24:25] 7 [08:24:56] clientTRIds helps clients track requests (like in epp) [08:25:14] add route group ( slide 8) [08:26:56] activate public Id (slide 9) [08:29:34] brian, sumath: need to encode a prefix, not just a range [08:30:48] hadriel on the mic [08:32:38] open numbering plan / ranges need to be more clearly specified. [08:33:55] next slides: commands to check if provisioning is correct, this is not the resolution protocol [08:36:27] sumanth: route group or route name? [08:41:18] jfm: asks for more feedback on the list [08:44:51] discussion on how to make the design team work more transparent [08:46:30] weekly meeting minutes should not be hard -- w3c does this for all meetings [08:50:15] Gonzalo leaves the room [08:51:16] next agenda item: transport spec [08:51:32] http://tools.ietf.org/agenda/78/slides/drinks-2.pptx [08:51:35] ken presenting [08:53:11] choice: thin SOAP/WSDL + http digest auth [08:54:56] nick russell presenting [08:55:08] http://tools.ietf.org/agenda/78/slides/drinks-3.ppt [08:57:22] operator standpoint: LTE with VoIP only, interconnection still mostly pstn based, need to to real IP peering for new services [08:57:54] slide 2: challenges [08:58:32] private interconnects, private networks like IPX [09:00:25] how can we handle user-owned domain name in a carrier-setting? [09:00:52] slide 5 [09:03:45] need a gloabl solution, not national solutions [09:04:34] slide 7 [09:05:31] Fabricio Tamusiunas joins the room [09:06:25] alex: know other sdos know what drinks is doing? [09:07:19] hanno, remarks from the 3gpp side [09:07:26] suzukisn joins the room [09:08:13] 3gpp contributors need to pull in IETF docs into 3gpp [09:08:33] hadriel atthe mic [09:11:05] problem of shared customer domain across multiple operators is hard [09:15:21] hadriel, jon-> worthwhile subject for the ietf. tricky to sell, go for a BOF [09:15:34] next agenda item: gSPID [09:16:15] what is output of the LUF: [09:16:38] domain ? administrative domain? numerical? [09:16:48] probably to URN [09:17:36] perhaps use IANA-registrered enterprise IDs [09:18:07] juan-francois on the mic [09:20:47] hadriel: perhaps a new URI-scheme? spid:user@ID ? [09:22:21] hadriel: do we need user@SPid, versus just the spid [09:23:54] jfm: define it here to save time [09:24:18] jon: uri-scheme takes a long time [09:24:52] nicc: reuse existing identifier schemes? [09:29:31] urn is self-typing [09:31:44] Jean-Francois M. leaves the room: Computer went to sleep [09:31:44] AOB [09:32:10] Fabricio Tamusiunas leaves the room [09:32:21] how do we proceed without speermint? [09:33:02] suzukisn leaves the room [09:33:27] jon: no problem, drinks can do stuff, or punt issues to dispatch [09:33:32] as0-d91k leaves the room [09:33:57] jgunn leaves the room [09:39:12] Otmar Lendl leaves the room: Computer went to sleep [09:39:52] tomkri leaves the room: Replaced by new connection [09:58:49] suzukisn joins the room [10:41:37] mlm.michael.miller leaves the room [11:34:19] suzukisn leaves the room [12:36:10] Fabricio Tamusiunas joins the room [12:36:40] Fabricio Tamusiunas leaves the room [13:31:41] burn leaves the room [15:11:13] Jean-Francois M. joins the room [21:06:20] Jean-Francois M. leaves the room [22:05:11] Jean-Francois M. joins the room [22:33:51] Jean-Francois M. leaves the room