[08:06:50] --- slavitch has joined
[09:51:17] --- nm has joined
[09:52:54] --- Esemwy has joined
[09:55:22] --- ft has joined
[09:59:29] --- newcat has joined
[10:00:11] --- dumdidum has joined
[10:01:36] --- wouter has joined
[10:01:56] --- newcat has left: Lost connection
[10:02:01] --- sunakichiwide has joined
[10:02:28] --- shidaschubert has joined
[10:02:47] --- agallant has joined
[10:02:50] --- newcat has joined
[10:02:52] --- csp has joined
[10:02:52] --- spencerdawkins has joined
[10:02:55] --- cullenfluffyjennings@gmail.com has joined
[10:03:56] --- tfroment has joined
[10:04:04] --- ekr has joined
[10:04:16] --- gas has joined
[10:04:47] --- tga73 has joined
[10:05:17] --- inet6num@jabber.org has joined
[10:05:21] --- jason.fischl has joined
[10:06:04] <spencerdawkins> are we in the right room? No jabber scribing yet?
[10:06:08] --- geir_egeland has joined
[10:06:19] <cullenfluffyjennings@gmail.com> yes - I think this is the right room
[10:06:28] <cullenfluffyjennings@gmail.com> at least all the other logical names did not work
[10:06:54] --- LouBerger has joined
[10:07:25] <spencerdawkins> thanks!
[10:07:25] --- cgn has joined
[10:07:35] <ft> does anyone have audio from Chantilly West? I can't hear anything on channel 6, but channel 7 is working :(
[10:07:40] --- nm has left: Replaced by new connection
[10:07:40] --- nm has joined
[10:07:40] --- nm has left
[10:08:00] <spencerdawkins> may also need to say what server (since the IETF is now using rooms.jabber.ietf.org)...
[10:08:23] <agallant> there is/was audio, but the signal level is very low, and has been all week
[10:08:46] <agallant> (on this channel, from this room)
[10:08:54] <inet6num@jabber.org> audiocast: ch6 (p2p-sip) is pretty bad.
[10:08:58] --- Esemwy has left
[10:09:09] --- miki has joined
[10:09:50] <ft> I got audio now, but with lots of stutter
[10:11:43] --- dwillis has joined
[10:12:22] --- javier has joined
[10:12:41] <spencerdawkins> could jabber scribe be on the old server?
[10:13:07] --- slavitch has left: Lost connection
[10:13:42] --- cabo--tzi--org@jabber.org has joined
[10:14:17] --- tianlinyi has joined
[10:15:21] --- dmalas has joined
[10:16:34] --- bhoeneis has joined
[10:16:42] --- LouBerger has left
[10:17:19] <spencerdawkins> * 1000 Discussion of Use Cases lead: Bruce B. Lowekamp read: draft-bryan-sipping-p2p-usecases-00 Have beein discussing on mailing list for several months, probably not complete (so please send text, preferably XML) Some scenarios may also work in client-server (still OK for p2p) Scenarios - global Internet, private networks, security, and maybe limited connectivity (ad hoc, disaster response) Global VoIP networks may be commercial or open. Commercial networks may include "super-peers" (but this is not Skype's architecture). Rohan - is a global network with no central authority in scope? Need to work with existing infrastructure (NATs, firewalls, etc.). "Scale unimportant", but this is mis-stated (want to work globally and on small networks) Idea that p2p-sip may be a failover mechanism
[10:17:40] <spencerdawkins> not just voice - also consumer electronics
[10:17:58] <spencerdawkins> PDAs, headsets, etc all bluetooth-connected
[10:18:06] --- nm has joined
[10:18:48] <spencerdawkins> could take pictures of receipts and send them to everyone at the table for expense forms - not voice apps, but leverage SIP infrastructure
[10:18:48] --- LouBerger has joined
[10:19:07] <spencerdawkins> "Access to ports blocked for 'security'"
[10:19:19] <spencerdawkins> could support anonymous communication
[10:19:33] <bhoeneis> can anybody open the slides on http://www3.ietf.org/proceedings/06mar/slides/p2p-sip-1.pdf
[10:19:37] <spencerdawkins> persecutions, avoid threats of violence
[10:19:40] <dwillis> working on it
[10:19:41] --- enrico has joined
[10:19:48] <bhoeneis> I am getting an error while opening
[10:20:05] --- mcfadden has joined
[10:20:25] <dwillis> server appears to be returning garbage, I'm working on it
[10:20:28] <spencerdawkins> small orgs - want private communications, regardless of member locations, external provider unacceptable
[10:20:53] --- plivesey has joined
[10:21:20] <spencerdawkins> limited connectivity - people with 770s at bars, developing world, "useful connectivity" - globally when available, locally when not
[10:21:22] <tianlinyi> to bhoeneis: failed to open
[10:21:23] --- Tony has joined
[10:21:24] <mcfadden> The PDF file does appear to be broken.
[10:21:25] <spencerdawkins> first responder
[10:21:49] <spencerdawkins> most first responders going all-digital already
[10:23:22] --- newcat has left: Lost connection
[10:23:56] <spencerdawkins> comment - these use cases are general, not p2p-SIP
[10:24:06] --- cullenfluffyjennings@gmail.com has left: Logged out
[10:24:09] --- newcat has joined
[10:24:49] <spencerdawkins> comment - geographical distribution of servers isn't p2p-sip-only - thinking about how to do this cheaply
[10:25:23] <spencerdawkins> rohan - small business use cases are disingenuous - don't think small companies will pay for them
[10:25:52] <spencerdawkins> servers aren't expensive - requirement is simple and cheap solutions
[10:26:17] <spencerdawkins> henry - "machines that are heavier than air will never fly" (chuckles)
[10:26:36] <spencerdawkins> can we get this guy closer to the mike?
[10:26:53] --- saverio.niccolini has joined
[10:27:00] <dwillis> which guy spencer?
[10:27:07] <spencerdawkins> greg daley - have housemate in accounting, on the road, don't have access to infrastructure
[10:27:14] <spencerdawkins> speakers at mike
[10:27:48] <spencerdawkins> this guy!
[10:27:56] <spencerdawkins> also a name would be good...
[10:27:57] <dwillis> slides should work now. server was trying to convert my pdf from ppt to pdf . . .
[10:28:31] <saverio.niccolini> i'm trying it now...
[10:28:36] --- audet.francois has joined
[10:28:43] <spencerdawkins> 1030 Discussion of Requirements lead: Henning Schulzrinne read: draft-basset-sipping-p2preq-00
[10:28:44] <agallant> plea to speakers: PLEASE SPEAK UP!
[10:29:30] <spencerdawkins> will focus on security requirements
[10:29:55] <spencerdawkins> defines terminology (need to extend this as we go, but we have many different backgrounds and terminologies)
[10:30:16] <spencerdawkins> synonymous with overlay networks (in this room)
[10:30:52] <spencerdawkins> basic idea is that resources of participants are shared to provide services - computation, bandwidth, and storage
[10:31:18] <spencerdawkins> may use limited centralized resources (think "1968 peer-to-peer networks")
[10:31:32] <tfroment> Is collaborative work, file sharing, etc... a use case for p2p sip?
[10:31:50] <spencerdawkins> current "p2p" systems are often central servers plus super-peers plus users
[10:32:39] <spencerdawkins> potential characteristics - good scalabilty, reduced management costs, reduced deployment costs, and something else (see slides)
[10:33:21] <spencerdawkins> distributed hash tables (DHT) mapping service that maps a string onto another string
[10:33:37] <spencerdawkins> "key and value mapping"
[10:34:58] <spencerdawkins> may have two-layer dht model
[10:35:23] <spencerdawkins> MUST support basic voice, videa, interactive text, may support other applications (presence, etc.)
[10:36:04] <spencerdawkins> distributed resources (location service, traversal servers, voicemail, address book, configuration storage)
[10:36:25] <spencerdawkins> also a generic mechanism to provide extensible mechanism
[10:36:34] <spencerdawkins> SIP is "close to" P2P already
[10:36:55] <spencerdawkins> proxies not mandatory, can be distributed
[10:38:58] <spencerdawkins> want to reuse as many needed protocols as possible
[10:39:04] <spencerdawkins> there is no ideal dht
[10:39:44] <spencerdawkins> want fast lookup, but also want good insert/deletion characteristics
[10:40:42] <spencerdawkins> cullen - DHTs don't interoperate, do they? (No) - Is the DHT something we are standardizing? (No)
[10:41:23] <spencerdawkins> DHT is still active research area. Client may be able to ignore DHT if external
[10:42:44] --- enrico has left
[10:43:19] --- mnvakil has joined
[10:43:43] <spencerdawkins> rohan - aren't we supposed to be talking about requirements?
[10:44:01] --- enrico has joined
[10:44:05] <spencerdawkins> NAT traversal is "SHOULD" - dominates server costs
[10:44:41] <spencerdawkins> should not introduce significant delays
[10:44:49] <spencerdawkins> DHTs aren't good at this today
[10:45:01] <spencerdawkins> Should support UDP and TCP
[10:45:45] <spencerdawkins> Should scale to Internet level, but must continue to function during church
[10:46:21] <spencerdawkins> should support both centralized and non-centralized naming models
[10:47:02] <spencerdawkins> Now have scenario with "evil server"
[10:47:18] <spencerdawkins> byzantine failure models
[10:47:49] <spencerdawkins> "at least 2/3s of systems are trustable"
[10:48:16] <spencerdawkins> rohan - separate out DOS attacks from other security considerations
[10:48:32] --- javier has left
[10:48:55] <spencerdawkins> need to protect against "mole invasion" (not just DOS attacks, etc.)
[10:50:28] --- cullenfluffyjennings@gmail.com has joined
[10:50:43] <spencerdawkins> "evil" nodes may not want to contribute resources, may want to eavesdrop, may want to actively attack and prevent communications
[10:51:37] <spencerdawkins> want to discourage nosy peers, but key exchange is hard
[10:51:51] <spencerdawkins> issue is MITM attacks
[10:52:32] <spencerdawkins> could build small-scale, built-in, or generic/external DHTs
[10:52:42] <spencerdawkins> can someone else type for a minute?
[10:53:11] --- dwillis has left
[10:53:17] --- enrico has left: Replaced by new connection
[10:53:42] --- enrico has joined
[10:54:03] <newcat> Regulatory issue?
[10:54:16] <newcat> e.g. 911
[10:54:19] --- cullenfluffyjennings@gmail.com has left: Logged out
[10:57:43] <spencerdawkins> back -
[10:57:47] <newcat> Having multiple peer with different funcationalities
[10:58:02] --- plivesey has left
[10:58:13] <spencerdawkins> henning's model - once you get to a resource it should work like 3261
[10:58:25] <spencerdawkins> maybe a subset, but that's the model
[10:58:35] <spencerdawkins> can we agree on these minimal services?
[10:58:54] <spencerdawkins> my question - does "minimal services" mean "core sip"?
[10:59:06] <spencerdawkins> dean - may be clearer after charter discussion
[11:00:12] --- aki has joined
[11:00:56] --- wouter has left
[11:01:12] <spencerdawkins> what about v4/v6 interworking? Henning hasn't thought about this
[11:01:24] <spencerdawkins> this is actually a friday thought
[11:01:46] <spencerdawkins> cullen - v4/v6 is OK for requirement, not just discussion on "how"
[11:02:46] <spencerdawkins> none of these requirements prevent service that you decide to trust
[11:03:01] --- Tsubasa has joined
[11:03:04] <spencerdawkins> we con't have selection requirements for search engines, either (not an IETF thing)
[11:03:35] --- lixia has joined
[11:03:41] <spencerdawkins> reputation services also required for peer-to-peer, but out of scope for now
[11:04:22] <spencerdawkins> rohan - requirement that any SIP node can send a message to a p2p-sip node? surprised that this isn't mentioned
[11:04:55] <spencerdawkins> it is not a requirement to have an overlay network
[11:06:08] <spencerdawkins> 0935 Scope, Draft Charter, and Organization of WG lead: Dean Willis
[11:06:32] <spencerdawkins> started with a very large scope, need a much smaller scope to charter a working group
[11:06:46] <spencerdawkins> have been discussing draft charter for about three weeks
[11:07:20] <spencerdawkins> this is not a replacement for DNS, not a peer-to-peer VoIP, only p2p-sip
[11:07:56] <spencerdawkins> want self-organizing/highly available proxy farm, not fixed hierarchy of IMS systems
[11:08:04] --- nm has left
[11:08:23] <spencerdawkins> topologies with ephemeral relations (mesh, emergency response, battlefield)
[11:09:15] <spencerdawkins> recovery functions when servers fail or endpoints isolated by network partitioning
[11:09:59] <spencerdawkins> is "self-organizing" technical or social? Dean thinks "technical", but community is also interested in social self-organizing (this is EKR)
[11:10:43] <spencerdawkins> dean assumes "association groups" - p2p nets, overlays, federations... but these are all name spaces
[11:11:00] <spencerdawkins> may map to domain-level identifiers in the DNS
[11:11:22] <spencerdawkins> don't have to have DNS to make this work, but probably easiest way to do this
[11:11:49] --- miki has left
[11:12:12] <spencerdawkins> is RHS a DNS thing? If you use DNS, yes - don't have to be in DNS, unless you are using DNS to resolve URIs that point to you
[11:12:42] <spencerdawkins> david schwartz - name space is a lot smaller scope than previously discussed
[11:13:19] <spencerdawkins> how to distinguish between names? based on rhs, but this is 3261-normal
[11:14:03] --- cullenfluffyjennings@gmail.com has joined
[11:14:37] <csp> mehdavi?
[11:14:41] <spencerdawkins> greg daley - may do entire lookup, not just lhs or rhs
[11:14:57] <spencerdawkins> looks close to what I'm hearing, csp!
[11:15:16] <spencerdawkins> lowcampf?
[11:15:36] --- miki has joined
[11:15:55] <spencerdawkins> Important thing on slide is "may" - can't assume DNS, CAs and collision resolution.
[11:16:24] --- nm has joined
[11:16:40] <spencerdawkins> there is a model with two overlay networks - how does alice in cluster A talk to bob in cluster B?
[11:17:11] <spencerdawkins> going-in position is that clusters use 3261 resolution between clusters
[11:17:49] <spencerdawkins> jonathan - need to be max interoperable with SIP - not closed Skype environment
[11:18:48] <spencerdawkins> normal SIP client would interact with adapters (actually, would be nice if this was explicitly direct interaction)
[11:19:25] <spencerdawkins> ??? nodes may detatch, but want to use the same name when I'm disconnected
[11:20:05] <spencerdawkins> dean - may be fundamental disconnect between this and using 3261 whether you are connected or not
[11:21:18] <spencerdawkins> This seems like a mismatch between community assumptions (Spencer's opinion)
[11:22:24] <spencerdawkins> mehdavi? within a p2p system, want to take my name and security problem can be solved by other means
[11:22:33] --- momose has joined
[11:23:02] <spencerdawkins> dean - this is the same disconnect as HIP (routability and identifier names)
[11:23:16] <spencerdawkins> cullen - back to the charter, please
[11:23:19] --- cwindenver has joined
[11:23:53] <spencerdawkins> assumption 2 - only difference from CS-SIP is mechanisms for location database
[11:24:22] <spencerdawkins> assumption 3 - some elements MAY be in DNS so they can be used for bootstrapping
[11:24:44] <spencerdawkins> use DNS if we have it available, use something else when we don't
[11:24:49] --- tianlinyi has left
[11:24:58] --- nagendra has joined
[11:25:16] <spencerdawkins> use RFC 3263 between clusters ("it's SIP")
[11:26:09] <spencerdawkins> assumption 5 - need at least one centralized identity mechanism
[11:26:28] <spencerdawkins> tasks slide -
[11:26:34] <spencerdawkins> rohan likes all of this
[11:26:48] <spencerdawkins> use zeroconf in all cases?
[11:27:01] --- aki has left
[11:27:40] <spencerdawkins> show generic operations with at least one binding (to prove they can be used for binding)
[11:28:35] <spencerdawkins> don't change turn, stun, etc. - take changes to other WG groups
[11:28:51] --- dumdidum has left: Replaced by new connection
[11:29:25] --- dumdidum has joined
[11:30:06] <spencerdawkins> section in existing charter that says "reuse existing stuff" - jonathan doesn't like "select" - lots of nat traversal mechanisms, need to use what SIP uses and fix it if it needs fixing
[11:30:56] <spencerdawkins> can someone type for a minute?
[11:32:16] --- newcat has left: Lost connection
[11:34:50] --- nm has left: Replaced by new connection
[11:34:51] --- nm has joined
[11:35:40] <spencerdawkins> back ---
[11:35:58] <spencerdawkins> agree on stun/turn/ice wording in draft charter
[11:36:05] <cullenfluffyjennings@gmail.com> Aki said the most important thing we should work on is the most expensive resource which is the media relays
[11:36:11] <spencerdawkins> thanks!
[11:36:28] <spencerdawkins> and it isn't in the current charter (did I get this right?)
[11:37:03] --- nagendra has left
[11:37:10] <spencerdawkins> mike s - media is solved, signaling is the issue, would use stun/ice/turn if they work for signaling
[11:37:24] <spencerdawkins> is that done in mmusic, sip, or p2p-sip?
[11:37:43] --- nm has left: Replaced by new connection
[11:37:43] --- nm has joined
[11:37:43] --- nm has left
[11:37:45] <spencerdawkins> dean - many people don't have a publically-addressible sip proxy in their models
[11:38:04] <spencerdawkins> can't use outbound, which requires sip proxy
[11:38:15] <spencerdawkins> use turn for binding?
[11:38:54] <spencerdawkins> jon - you're saying you agree with the text, right?
[11:39:03] <spencerdawkins> (sorry, that was jonathan)
[11:39:34] <spencerdawkins> we can wordsmith, not here
[11:39:35] --- nm has joined
[11:39:49] <spencerdawkins> jonathan agrees
[11:40:03] <spencerdawkins> with existing text
[11:40:24] <spencerdawkins> greg daley - like text, it's good, don't spend three years working on stuff like this
[11:40:36] <spencerdawkins> won't be any telecom left, let's get this done.
[11:40:56] <spencerdawkins> "signaling is just another medium" - write deltas in the draft
[11:41:11] <spencerdawkins> tim chown - don't want lots of baggage in architecture
[11:41:24] <spencerdawkins> include v4/v6 interworking in this section?
[11:41:43] <spencerdawkins> dean - put this back into mission parameters slide (that I'm trying to talk about)?
[11:42:01] <spencerdawkins> can't not do v4/v6 in today's world
[11:42:25] <spencerdawkins> tim - but APP open area meeting talked about this
[11:43:29] --- newcat has joined
[11:43:29] --- newcat has left: Lost connection
[11:43:34] --- newcat has joined
[11:44:05] <spencerdawkins> rohan - reinforcing NAT thing - objections to SIP-outbound as signaling mechanism - if you don't have l2 domain and don't have any infrastructure at all, am incredulous that we still need signaling??? incredulous ("without a public supernode")
[11:44:55] <spencerdawkins> jonathan - need to debunk "SIP outbound assumes public internet" - need a physical point of IP interconnect, if you don't have that, you're not sending packets
[11:45:18] <spencerdawkins> SIP outbound requires some point everyone can reach - that's the requirement
[11:45:47] <spencerdawkins> cullen - understand basic use case, lots of discussion that we need to defer
[11:46:16] <spencerdawkins> working across complete symmetric NAT network is impossible
[11:46:19] --- cullenfluffyjennings@gmail.com has left: Logged out
[11:46:38] --- saverio.niccolini has left
[11:47:25] <spencerdawkins> ekr - want to push back on two very different designs - you may or may not know what the backend is doing, and may or may not be part of it
[11:47:56] <spencerdawkins> is there another slide after this one? (spencer)
[11:48:10] --- saverio.niccolini has joined
[11:48:25] <newcat> Mission parameter slide
[11:48:38] --- saverio.niccolini has left
[11:49:09] <spencerdawkins> ekr - need architecture choices, not just design choices
[11:49:30] <spencerdawkins> dean - this is already on slide (bullet 2) - do we need two architectures?
[11:49:54] <spencerdawkins> ekr - if you can't pick between these two architectures, you're not ready to charter
[11:50:49] <spencerdawkins> jonathan - architecture is definition of components and how they relate to protocols - need to know what you're trying to do before you can say you have an architecture
[11:51:04] --- cwindenver has left
[11:51:17] <spencerdawkins> may have some architectures that leave no work to do, others that require work before they function
[11:51:29] <spencerdawkins> dean - intent is to cover both approaches
[11:52:29] --- csp has left
[11:52:56] --- cullenfluffyjennings@gmail.com has joined
[11:53:42] --- agallant has left
[11:54:07] --- enrico has left
[11:54:17] <spencerdawkins> henning - having a problem reading the charter text without having read the mailing list
[11:55:00] <spencerdawkins> ??? this is all great stuff, but should we explicitly say "plug and play" or something similar? Don't have this in the charter (group seemed to node)
[11:55:15] <spencerdawkins> not asking for invention here
[11:55:26] <spencerdawkins> jon the AD - getting increasingly worried
[11:55:45] <spencerdawkins> think I understand charter for achievable work
[11:56:01] --- plivesey has joined
[11:56:17] <spencerdawkins> is this going to be formed by broad set of design choices? sounds like a research group
[11:57:21] <spencerdawkins> first two tasks look like roadblocks to 3 and 5 tasks - please tell me I'm wrong
[11:57:53] <spencerdawkins> henning - we have at least three working prototypes with at least two models covered
[11:58:23] <spencerdawkins> charter doesn't reflect the community's level of experience in this area
[12:01:16] --- dumdidum has left: Replaced by new connection
[12:01:25] --- dumdidum has joined
[12:01:44] --- inet6num@jabber.org has left
[12:02:43] <spencerdawkins> jon - scenarios and architecture is a quagmire, strongly suggest WG focus on retrieving location and updating it securely
[12:03:41] <spencerdawkins> greg - there is already an architecture draft, we'll look at it on Friday, we can talk about this then
[12:04:11] <spencerdawkins> is "defining services" in scope for this charter? at least rendezvous? configuration?
[12:04:52] --- dmalas has left
[12:06:41] --- eoin.mcleod has joined
[12:08:04] <spencerdawkins> jonathan - if call forward no answer is a service that we're talking about, does it have to work the same way in this environment?
[12:08:13] --- nm has left
[12:08:23] --- geir_egeland has left
[12:08:36] <spencerdawkins> dean - not the phone network, doesn't have the same service model and doesn't have the same user expectations
[12:08:49] <spencerdawkins> "not the class 5 switch in the sky"
[12:09:01] --- tfroment has left: Replaced by new connection
[12:09:27] <spencerdawkins> my voicemail inserts into the network at a lower priority level, and just answers when I don't.
[12:09:43] --- jason.fischl has left
[12:09:44] --- danwing has joined
[12:09:57] <spencerdawkins> there's no mandatory-to-implement for observing queue value?
[12:10:33] <spencerdawkins> scribe - need to walk before we can run, need to solve any problem at all before we solve all problems
[12:11:18] --- nm has joined
[12:12:28] --- cullenfluffyjennings@gmail.com has left: Logged out
[12:12:32] <spencerdawkins> dean - could have an application server that registers for you (peer to peer serving csf, just shoot me)
[12:13:27] --- Tsubasa has left
[12:13:39] <spencerdawkins> jon - we're complicated enough already now
[12:13:54] --- tfroment has joined
[12:14:15] --- dumdidum has left: Replaced by new connection
[12:14:30] <spencerdawkins> dean - we may want to make a list of things that are in scope, but isn't on our current list
[12:15:07] <spencerdawkins> dean doesn't know what "using p2psip mechanisms to management groups of distributed media relays" means on "excluded" slide - this came in from the list
[12:15:49] --- dumdidum has joined
[12:15:58] <spencerdawkins> jon - do people think taking on both media and signaling in the first version of the charter is a good thing?
[12:16:29] <spencerdawkins> ??? media part is well understood, signaling and overlap maintenance is the hard part.
[12:17:04] <spencerdawkins> ??? would like to see a system that doesn't rely on centralized media relays or stun servers
[12:17:08] <spencerdawkins> (this was aki)
[12:17:20] <spencerdawkins> dean - complete system or components?
[12:17:31] <spencerdawkins> jon - not going to standardize a system here
[12:17:50] <spencerdawkins> aki - cut down to what implementers implement?
[12:18:20] <spencerdawkins> learn from simple - didn't have publish in first version of charter
[12:19:29] <spencerdawkins> jonathan - doing nat traversal later is worse than doing security later
[12:20:05] <spencerdawkins> cullen - if we do media relays, are we also picking the right one? that's a tricky problem
[12:20:36] <spencerdawkins> greg - there are people who have time for this - want to make sure nat traversal for p2p sip
[12:20:43] <spencerdawkins> can't see any other way
[12:23:28] <spencerdawkins> ??? don't spend ten years on architecture, but spend a little time on it before someone has to implement from interface specs with no semantics
[12:23:55] <spencerdawkins> "you go figure it out:"
[12:24:08] --- nm has left: Replaced by new connection
[12:24:08] --- nm has joined
[12:24:08] --- nm has left
[12:24:17] --- nm has joined
[12:24:21] <spencerdawkins> dean - historically, ietf doesn't publish architectures
[12:24:39] <spencerdawkins> ??? can provide guidance without standardizing an architecture
[12:25:54] <spencerdawkins> out of scopes - non-SIP, replacing SIP, research-type problems (including p2p replacement for SIP)
[12:26:04] <spencerdawkins> must cooperate with other working groups
[12:26:22] <spencerdawkins> avoid extension or changes whenever possible
[12:27:15] --- cgn has left
[12:27:57] --- Tony has left
[12:29:21] --- audet.francois has left: Logged out
[12:30:19] --- dumdidum has left: Replaced by new connection
[12:30:28] --- cgn has joined
[12:30:50] --- dumdidum has joined
[12:32:13] --- sunakichiwide has left
[12:32:29] --- tfroment has left: Disconnected
[12:32:32] <spencerdawkins> sitting down from the mike, but I am concerned that we are saying that we must be able to produce a proposed standard because we have too much commercial pressure to figure out what the problem is and/or experiment?
[12:32:38] <spencerdawkins> that's not good.
[12:33:43] --- danwing has left: Replaced by new connection
[12:34:15] --- spencerdawkins has left
[12:34:22] --- plivesey has left
[12:34:29] --- tga73 has left
[12:34:34] --- lixia has left
[12:35:05] --- nm has left
[12:35:48] --- newcat has left
[12:35:48] --- cgn has left: Logged out
[12:35:49] --- gas has left
[12:36:01] --- momose has left
[12:36:08] --- eoin.mcleod has left
[12:36:19] --- bhoeneis has left
[12:42:57] --- mnvakil has left
[12:51:42] --- cabo--tzi--org@jabber.org has left
[12:52:32] --- mcfadden has left
[12:53:34] --- dumdidum has left
[13:03:23] --- LouBerger has left
[13:10:26] --- spencerdawkins has joined
[13:33:09] --- ft has left
[13:50:13] --- shidaschubert has left
[13:52:15] --- ekr has left
[13:53:23] --- spencerdawkins has left
[13:59:34] --- spencerdawkins has joined
[14:04:04] --- miki has left
[14:04:28] --- spencerdawkins has left
[15:53:20] --- dumdidum has joined
[15:53:50] --- dumdidum has left
[17:22:21] --- sunakichiwide has joined
[17:22:31] --- sunakichiwide has left