[00:54:07] --- jm has joined
[00:54:25] --- jm has left
[01:04:49] --- dlpartain has joined
[02:35:18] --- gih has joined
[02:35:24] --- gih has left
[03:02:16] --- dlpartain has left
[03:17:42] --- vlevigneron has joined
[03:18:56] --- vlevigneron has left
[03:35:16] --- klensin-ietf has joined
[03:54:12] --- Lisa D has joined
[03:57:05] --- dcrocker has joined
[04:17:57] --- smb has joined
[04:18:21] --- omarjan has joined
[04:20:58] --- omarjan has left: Disconnected
[04:21:48] --- vlevigneron has joined
[04:23:05] --- jakob has joined
[04:23:47] --- vlevigneron has left
[04:24:23] --- vlevigneron has joined
[04:24:51] --- javier has joined
[04:26:45] --- lha has joined
[04:26:57] --- sakai has joined
[04:27:24] --- omarjan has joined
[04:27:30] --- Ted Faber has joined
[04:27:34] --- anewton has joined
[04:27:34] --- shep has joined
[04:27:38] --- Ole has joined
[04:27:53] <anewton> leslie.
[04:27:57] --- masahiro has joined
[04:28:05] <anewton> Introductions. agenda.
[04:28:06] --- hardie has joined
[04:28:36] --- tskj has joined
[04:28:51] --- mstjohns has joined
[04:28:52] <anewton> Eric Nordmark.
[04:29:35] --- bkhabs has joined
[04:29:43] --- hta has joined
[04:29:47] * anewton quizzing the powers that be on location of the slides... hold on.
[04:30:07] <hta> slides: http://www.alvestrand.no/ietf/korea/
[04:30:31] <hta> pdf & ppt (where relevant)
[04:30:39] <anewton> multi-homing. connect site to multiple ISPs, improved failure resilience, load balancing, better quality connectivity
[04:30:46] --- resnick has joined
[04:30:51] --- hartmans has joined
[04:31:27] --- leslie has joined
[04:31:27] <anewton> today's internet.
[04:31:34] --- tonyhansen has joined
[04:31:45] --- warlord has joined
[04:31:53] <anewton> multihoming todya. provider independent addressing, routes can't be aggregated by ISPs
[04:32:57] <anewton> v6 multihoming today
[04:33:00] --- sleinen has joined
[04:33:16] <anewton> could use ipv4 approach. or multiple addresses per host
[04:33:31] --- kdz has joined
[04:33:45] --- ludomp has joined
[04:33:51] --- galvinjamesm has joined
[04:34:11] <anewton> multiple address per host. resilience provided in the application layer. needs address selection. ingress filtering issue.
[04:35:25] <anewton> rfc 3484 good start
[04:36:28] --- tuy has joined
[04:36:38] --- Netrangerrr has joined
[04:36:45] <anewton> separate identifiers and locators. ip addresses used for identifying purposes, locating purposes
[04:37:07] --- jhutz has joined
[04:37:19] <anewton> intentionally designed overloading. time to seperate into identifiers and locators? doesn't imply new ID name space
[04:37:19] --- rei has joined
[04:37:42] --- kjd has joined
[04:38:36] <anewton> dns, apps, transports operate on IDs. ip operates on locators. new layer on hosts map between identifiers and locators. need protocol to setup mapping on hosts
[04:39:39] <anewton> new layer of indirection. transparent to (most) applications. possible to lookup identifier to find locators? exact location of the shim?
[04:40:01] --- kjd has left
[04:40:05] --- BP has joined
[04:40:23] --- kjd has joined
[04:41:45] <anewton> new namespace or not? could add the layer of indeirection wihtou a new name space. new identifers could be long-term stable. what benefist do they have?
[04:42:42] --- avri has joined
[04:43:32] <anewton> what could new ID look like? allocate using managed hierarchy? our self-allocate?
[04:43:39] <anewton> s/our/or/
[04:44:33] --- arifumi has joined
[04:45:12] --- sommerfeld has joined
[04:45:28] <anewton> the mapping system. allow applications to lookup identieifer to find locators. we only know how to make this scale for hierarchical name spaces. can it be made to scale in the slf-mangaged case?
[04:46:35] <anewton> connection rehoming. able to change locators. needed wheter or not a new ID name space is used. need a protocol to "rehome" the traffic. multi6 WG has many proposals for this.
[04:47:09] <anewton> ingress filtering. isps can be told of all the address prefixes. different approaches.
[04:47:48] --- Melinda has joined
[04:49:44] --- agaton has joined
[04:49:54] <anewton> path/locator selection. the shim layer needs to select the source and destination locators to use. this is largely independenet of new ID namespace, connection rehoming. is there experience with sctp which we can leverage?
[04:50:11] --- ggm has joined
[04:50:57] <anewton> path/locator failure detection. time constraints are application dependenet. transport protocols can provide hints. role of the routing system? shim layer could echange heratbeats end-to-end.
[04:53:09] --- rlbob has joined
[04:53:10] <anewton> time line/ambition. 1. something short term to make hosts with multiple prefixes work well. 2. hide multiple IP addresses without a new name space. 3. introduce a new name space
[04:53:17] <anewton> might need 1 for 2 or 3
[04:53:40] --- MattMathis has joined
[04:54:31] --- randy_g has joined
[04:54:33] <anewton> security ambition. the new multihoming protocols should not make the Internet less secure than today. as things like dns and bgp are secured, we don't want the multihoming protocols to become the weakest link.
[04:54:40] --- Loughney has joined
[04:55:23] <anewton> security thread: redirection attact. connection rehoming implies changeing locators. if ipsec, tls, etc used, attack limited to selective dinaial-of-service attacks
[04:57:05] <anewton> secirty threat. 3rd party flooding. today flooding capability of attacker is limited by its bandwidth. being able to specify the locator of a 3rd party changes this.
[04:59:12] <anewton> security approaches. don't depend on some not-yet-deployed infrastructure, such as global pki. different approaches have been suggested. security implies more work for the hosts.
[05:00:16] --- Valissystem has joined
[05:01:05] <anewton> what about mobile ip? mobile ipv6 work a baseline for security understanding. effectively has a stable locator (home address) and a temporary, smaller latency locator (care-of address). this can be generalized for multihoming. scaling: solution for sites might not scale to host
[05:01:53] <anewton> current status. MULTI6 WG HIP WG
[05:02:05] <anewton> proposed HIP-related research group.
[05:02:24] --- tomphelan has joined
[05:03:22] <anewton> how can you help? how useful would a new ID space be to applications? what properties would it need to satisfy? what are the issues if fqdns and / or sets of locators are the indentifiers.
[05:03:48] <anewton> can bgp/igps provide faster hints that a locator might not work? how to perform e2e failure?
[05:04:14] <anewton> credits. mike o'dell's 8+8... didn't get the whole slide
[05:04:17] <anewton> charle perkins.
[05:04:35] <anewton> site multihoming, nemo wg have related work?
[05:04:39] <anewton> eric.
[05:04:53] <tuy> credits mention SCTP too
[05:04:55] --- akihisau has joined
[05:05:06] <anewton> don't think it is the place where it is invented. other work going on looking at multihoming
[05:05:08] <anewton> charlie.
[05:05:20] <anewton> network moving from one place to another is like moving from isp to another
[05:05:37] <anewton> is hip to help with site multihoming, need hierarchical id structure?
[05:05:40] <anewton> eric.
[05:05:55] <anewton> not an expert.
[05:05:59] <anewton> charlie.
[05:06:20] <anewton> need way to avoid mapping to be bottleneck.
[05:06:26] --- leslie has left: Replaced by new connection
[05:06:26] <anewton> ??.
[05:07:27] <anewton> discussion of multi6 group is not an issue, because it is not related.
[05:07:37] * anewton I'm having a really hard time with this
[05:07:39] <hta> ohta-san: multihoming and mobility are not related
[05:07:42] --- leslie has joined
[05:07:59] * anewton thanks
[05:08:09] <hta> ohta: the other example is threat analysis
[05:08:24] <anewton> redirection attack already available today with tcp
[05:08:25] <hta> ohta: redirection attack is already there today with today's tcp
[05:08:58] <anewton> eric.
[05:09:14] <anewton> need to discuss details in wg, purpose of here is to inform community
[05:09:31] --- marushin has joined
[05:09:34] <anewton> ohta.
[05:09:50] <anewton> it is fair for me to discuss this here because you mentioned it here.
[05:10:04] <anewton> eric.
[05:10:19] <anewton> this has been going on for 10 - 30 years. this discussion goes about a long time
[05:10:28] <anewton> geoff.
[05:10:30] <hta> I reminded ohta that there were people behind him in line......
[05:10:53] --- ft has joined
[05:10:56] <anewton> structured/unstructured ID spaces. birthday problem. is very real.
[05:11:16] --- saturn has joined
[05:11:21] <anewton> issue of resolution of IDs, if structured it is easy to see a lookup mechanism
[05:11:28] <hardie> specific to opportunistic IDs, I believe, in Geoff's comments
[05:11:36] <anewton> doesn't birthday problem happen when they are put together.
[05:11:49] <anewton> erc.
[05:12:19] <anewton> need to look at that deeper in this space. lookup system will have collisions if everybody has an entry
[05:12:27] <anewton> geoff.
[05:12:40] <anewton> gives numbers to back up conclusion.
[05:12:48] <anewton> roham.
[05:12:55] <hardie> Rohan
[05:12:56] <anewton> rohan.
[05:13:01] --- Joe Touch has joined
[05:13:09] <anewton> big giant NAT, that's what it looked like.
[05:13:20] <anewton> look at implications on peer to peer applications
[05:13:25] <anewton> eric.
[05:13:59] <anewton> refers to slide, id/locator gets rewritten both ways, looks like a NAT
[05:14:17] <anewton> system designed to do it without impact
[05:14:22] <anewton> rohan.
[05:14:27] <anewton> dependent on topology
[05:14:29] <anewton> eric.
[05:14:35] <anewton> done at every host, so no
[05:14:46] <anewton> dave crocker.
[05:14:52] <anewton> thanks eric.
[05:14:54] --- Valissystem is now known as jmc
[05:15:01] <Loughney> just as an FYI, slides are here: http://www.alvestrand.no/ietf/korea/
[05:15:19] <anewton> mobility and multihoming look the same
[05:15:58] <anewton> need to be driven by requirements, none offered. this is an additional need. your slides assume new ID space. not necessarily wrong.
[05:16:05] --- mccreary has joined
[05:16:40] <anewton> diff between ID during initial phase vs. one that might be used in the assocation.
[05:16:41] --- Bill has joined
[05:16:47] --- ignas has joined
[05:17:00] --- galvinjamesm has left: Replaced by new connection
[05:17:01] --- galvinjamesm has joined
[05:17:01] --- galvinjamesm has left
[05:17:08] --- galvinjamesm has joined
[05:17:16] <anewton> might want to start with mobility and add in multihoming
[05:17:20] <anewton> eric.
[05:17:35] <anewton> depends on time scales and need for renumbering.
[05:17:48] <anewton> dave.
[05:18:22] --- howard has joined
[05:18:24] <anewton> the problem is site multihoming where host multihoming is ??.. the view may be the other way
[05:18:28] <anewton> eric.
[05:18:33] <anewton> not assuming new name space.
[05:18:47] <anewton> dave.
[05:18:52] --- mengwong has joined
[05:19:03] --- kivinen has joined
[05:19:04] <anewton> it is tough, did say you did it, but there is that tendacy to happen
[05:19:09] <anewton> eric.
[05:19:34] <anewton> are we driving from requirements? what are the actual issues for applications?
[05:19:43] <anewton> is it something applications could leverage?
[05:20:04] <anewton> would be good if driven by people higher in stack.
[05:20:12] <anewton> allison mankin.
[05:20:28] <anewton> good idea to sketch non-requirements
[05:20:44] <anewton> consider pruning path selection based on path quality.
[05:20:58] --- howard has left
[05:21:23] <anewton> not an integral part of the problem
[05:21:24] --- howard218 has joined
[05:21:27] <anewton> eric.
[05:21:32] <anewton> agree
[05:21:32] --- saturn has left
[05:21:41] <anewton> allison.
[05:21:49] <anewton> some question about which layer on which function
[05:21:54] <anewton> kurtis linquist.
[05:22:16] <anewton> need input from people of application and transport.
[05:22:22] <anewton> in multi6
[05:22:49] <anewton> ohta.
[05:23:07] <anewton> proposing same thing, not surpizing.
[05:23:17] --- JonneSoininen has joined
[05:23:27] --- howard218 has left
[05:23:28] --- howard218 has joined
[05:23:28] --- howard218 has left
[05:23:42] <anewton> your approach for some but most applications using tcp.
[05:23:47] <anewton> but not beyond that.
[05:23:50] <anewton> eric.
[05:23:53] --- howard218 has joined
[05:23:55] --- tonyhansen has left: Disconnected
[05:23:59] <anewton> things are more sudtle than that.
[05:24:36] <anewton> don't think it affects what it looks like to the applications no matter where it occurs in the stack.
[05:24:40] <anewton> ohta.
[05:24:42] <anewton> no.
[05:24:51] --- tonyhansen has joined
[05:24:57] <anewton> your proposal belongs to transport layer or some connection layer.
[05:25:01] <anewton> leslie.
[05:25:06] <anewton> cutting mic lines now.
[05:25:09] <anewton> charlie perkins.
[05:25:24] --- Glenn Parsons has joined
[05:25:28] <anewton> apps using locators vs IDs
[05:25:37] <anewton> eric.
[05:26:01] <anewton> not about locators getting to the applications, but ids. don't need apps to select between locators
[05:26:14] <anewton> ??.
[05:26:32] <anewton> hip is routing overlay. good idea.
[05:26:45] <bkhabs> Joe Touch
[05:26:51] <anewton> side effect requires hosts to run some participation of routing protocol
[05:26:59] --- howard218 has left
[05:27:05] <anewton> not a new thing, just a natural consequence of what you're doing
[05:27:14] --- howard218 has joined
[05:27:24] <anewton> going down architectural path that is not sustainable.
[05:27:26] <anewton> eric.
[05:27:37] <anewton> poeple are scared because it looks like nat.
[05:27:39] --- ft has left: Lost connection
[05:27:39] --- agaton has left: Lost connection
[05:27:39] --- lha has left: Lost connection
[05:27:49] --- ft has joined
[05:27:50] --- jakob has left
[05:27:52] --- lha has joined
[05:27:53] <anewton> but it isn't that way.
[05:28:00] <anewton> joe.
[05:28:13] <anewton> just make sure it isn't a nat
[05:28:19] <anewton> ??.
[05:28:40] <sommerfeld> gabriel montenegro
[05:28:52] <anewton> work in ?? about bootstrapping.
[05:29:05] <ggm> new anchor points, homes, first Q is
[05:29:16] <ggm> make it even more ameanable to application in v6, the bootstrapping
[05:29:33] <ggm> erik. trying to get work done on details, what makes sense for security, where to anchor
[05:29:46] <ggm> hard to say if this type of thing could be a piece in the puzzle
[05:29:55] <ggm> worth doing the analysis.
[05:30:38] <ggm> look at all the work to be done . duplicate, or work in different ways, each doing stuff, not worrying about architecture. is the IETF structured to solve this?
[05:31:14] <ggm> Erik Q for next session. thinking how to leverage, boiling the ocean. whats relationship of this to mobile IP, link and never get done. elaborate what doing, make more uniform, makes a lot of sense.
[05:31:18] <ggm> coordinating, thats a challenge
[05:31:41] <ggm> Related to last Q. liked presentation a lot, didnt see here, difficulties here, cannot solve. in that case.. what is plan B?
[05:31:45] <ggm> Erik
[05:31:57] <ggm> depends what you mean by solve. kernel have to satisfy, everything else is 'desired'
[05:32:13] <ggm> want to be everything, failover in 3ms. constrain, engineering tradeoffs
[05:32:22] <ggm> Tim Shephard. nice presentation.
[05:32:45] <ggm> Fan of HIP stuff, following with much interest. problem of how to map ID, map to locater, has been problem
[05:33:40] <ggm> want to do referralls with ID, showstopper, find locater. tech dependencies in my mind, duck the Q and avoid it. perhaps there is some way out. referrals still done with locaters, security problem. or referrals include ID. like with humans. bundle of stuff. going to be a researchy Q for a while.
[05:33:43] <ggm> Other comment
[05:34:55] <ggm> what I've seen in multi6. lot of proposals floating around. good to see code. more importantly than run, read. build enough code, can come in and do a demo, read the code to pull off the demo. lot of value in seeing how its really done. done enough to do the demo (can be faked) and show code, can figure out whats going on. lot I'd find for my own education, to read the code, pull off these proposals, more than draftsm hard to understand, code easier to read, runs ..
[05:35:01] <ggm> Erik. agree prototyping is good.
[05:35:17] <ggm> also agree first point. referrals at app level differently, assume 128bit number we pass along
[05:35:25] <ggm> having people explore that sort of angle very useful.
[05:35:42] <ggm> Leslie. Thanks Very Much Erik
[05:35:48] <ggm> [applause]
[05:35:49] --- mattz has joined
[05:36:40] <ggm> harald. next step is admin restructure. have worked together on this for quite a while. Leslie is taking this part, I;m taking the next part.
[05:36:44] <ggm> Leslie.
[05:36:52] <ggm> adv. comm report talked about at last plenary session.
[05:36:56] --- kurtis has joined
[05:37:02] <ggm> what we're going over tonight is a status report, whats been done since then
[05:37:20] <ggm> the assumption is people here are largely familiar with the docs. didn't post updates.
[05:37:27] <ggm> not that much news.
[05:37:32] --- kdz has left
[05:37:44] <ggm> three docs harald and I put out as individual drafts before cuttoff.
[05:37:57] <ggm> proposal for mission statement. Harald to go over this in sep agenda item. name listed.
[05:38:12] --- kdz has joined
[05:38:14] <ggm> there was a doc describing motivation for admin restructure. big blob of text is summary (slide)
[05:38:29] <ggm> words put together carefully. appropriate to quote directly.
[05:38:53] <ggm> key items. following from adv. comm. work. find mechanism to 'load balance' resources and work, to meet needs of the admin side, getting IETF more time going forward.
[05:39:21] <ggm> third doc. actual proposal for administrative restructuring. really written at level of arch requirements.
[05:39:28] <ggm> not a specification. not enough details to call it one.
[05:40:43] <ggm> spec of actual concrete proposal would include stuff like draft bylaws. probably have, will have Qs are 'how to impl....' 'why doesnt doc spec zyx' -will come. but its
[05:40:46] <ggm> and architecture document
[05:40:54] --- narten has joined
[05:40:56] <ggm> by way of practical overview.
[05:41:17] --- Melinda has left: Disconnected
[05:41:24] <ggm> put together admin pieces so we can tackle some of the larger problems. eg better tools for tracking. IANA assignments. RFC-ed.
[05:41:27] --- lynn has joined
[05:41:43] <ggm> improve efficiency & effectiveness of tracking. weeks/months of document stuck, could have solved by a conversation 3 months ago.
[05:42:33] <ggm> then there is coordination. funds. work. accountability. last of all. and leslies fave: not done by IAB or IESG directly. currently stuff requires a lot of day to day direction. not why the nomcom appointed people.
[05:42:55] <ggm> status: have proposal docs. please send comments to authors. have been discussed in IAB IESG. largely behind these proposals. send comments to IAB/IESG
[05:42:58] <ggm> next steps
[05:43:32] <ggm> refine proposals. are internet drafts, open for comment. marching along, getting data with specific impl. proposals. working with ISOC. consider as 'running code'
[05:43:55] <ggm> status back to ietf@ietf.org list. I am getting zero feedback. so. thats where we've gone since adv. report at last plenary
[05:43:59] <ggm> comments/Questions?
[05:44:09] <ggm> [silence]
[05:44:17] <ggm> OK. Jonas. tried to go to right mike.
[05:44:20] --- smb has left
[05:44:26] --- mstjohns has left
[05:44:28] <ggm> wanted to ask. working on this right now, you and harald, or others too?
[05:44:30] <ggm> Leslie
[05:44:44] <ggm> me and harald, talking to lindsay, IAB IESG involved,but mostly us
[05:44:51] <ggm> Jonas had some comments
[05:45:02] <ggm> Leslie to authors, more general comments to IAB/IESG or to us, and we'll forward
[05:45:17] <ggm> Jonas other thing is said going to do project on this. how can contribute, help?
[05:45:21] --- azinin has joined
[05:45:23] <ggm> Leslie will become clear as project proposed.
[05:45:39] --- bonninjm has joined
[05:45:40] <ggm> Regios?
[05:45:50] <ggm> one Q. then help me figure how to go forward.
[05:45:53] <ggm> arch requirements doc.
[05:45:55] <ggm> Leslie yes
[05:46:21] <ggm> Chapter entitled board of admin review. seems to be requirement for board to be composed of IAB/IETF chair. and up to 3 other people. overloaded but setting up requirement for more work. seems ...
[05:46:38] <ggm> Leslie good concern. if text doesnt say, is trying to say 'appointed by...' not drawn from. important distinction.
[05:46:46] <ggm> Dave Crocker
[05:46:51] --- azinin has left
[05:47:25] <ggm> Following up on that. a group like that board, in, this kind of situation, needs to be pretty diverse, different views etc. having them appointed from such a small range of sources, also thought exofficio IAB/IETF chair
[05:47:33] --- masahiro has left: Disconnected
[05:47:35] <ggm> Leslie does say ex-off but also ..
[05:47:43] --- Melinda has joined
[05:47:46] --- Melinda has left: Replaced by new connection
[05:47:52] --- Melinda has joined
[05:47:55] <ggm> Dave 3 people derived from same community. not independant set of selection processes. need to look at drawing from some other source
[05:47:57] --- Melinda has left: Replaced by new connection
[05:47:59] <ggm> Larry.
[05:48:53] <ggm> sounds like wanted a bit of process on maintenance s/w to help support... then suprised by scope of it, committee of long standing, purpose to fund th ebudget to make the thing happen. starting large not small. maybe misreading it. how big a budget are you thinking of? how long? scope it. the project, in your mind how big. admin restructure not process fiddle
[05:49:20] <ggm> Leslie. doc clearer about suite of issues facing IETF now, take care of its administrative work. not proposing incorperat so we can make WG tools. WG tools, cross
[05:49:31] <ggm> org tracking is pragmatic, functionally hard to implement in given env.
[05:49:46] <ggm> Larry but could we START with a little bit of s/w and an admin group to .. start small.
[05:49:58] --- memo56 has joined
[05:49:58] <ggm> Leslie. going to be rude; have you read all four docs.
[05:50:04] <ggm> Ohta.
[05:50:42] <ggm> never read any of the drafts but I think can comment. most of us have freedom to leave IETF and go to other standards bodies. Concern mostly is on how easy to make proposal, some standard, and how quickly, if you fail to do so, many of us will leave.
[05:50:59] <ggm> Leslie: if I've understood your point, this is nice, but doesnt matter if process doesnt work
[05:51:29] <ggm> ?missed name? interested in tools, tracking. Q is how going to proceed, development, concentrate on some other org, or ..
[05:51:30] --- akihisau has left: Replaced by new connection
[05:51:39] --- akihisau has joined
[05:51:45] <BP> Name: Dan Forsberg
[05:51:55] <ggm> Leslie. proposal of a kind of project we could take on. specifically how, would be aligned with some work Harald is going to talk about. other efforts for restructuring ietf
[05:51:59] <ggm> Harald.
[05:52:01] --- hta has left
[05:52:17] --- sob has joined
[05:52:47] <ggm> one document furthest along, least dependencies on others is the IETF mission statement
[05:52:57] <ggm> org which needs a mission statement is in serious trouble [laughter]
[05:53:10] <ggm> I dont thnk the existance of the mission statement changes that.
[05:53:12] <memo56> Is there actually any reason against discussing the advisory docs in a maillist?
[05:53:22] <ggm> state principles ORG relies on.
[05:53:34] <ggm> be guideline/yardstick to see what we do helps, or hinders the mission
[05:53:49] <ggm> side beneit: stop the email stoprs about 'what is mission'
[05:54:19] --- JonneSoininen has left
[05:54:24] <ggm> short version: goal of the IETF is to make the Internet work
[05:54:38] <ggm> has one slide for each. structured in 3 sections
[05:54:49] <ggm> <the rest of you know what to do..>
[05:55:30] <ggm> particular part of making it work, that the IETF does, is its mission. hiqual, tech, relevant docs, influence the way people use/design & manage, to make it better
[05:55:53] <ggm> including IP over avian carrier, and .sex considered harmful
[05:56:14] <ggm> open process, anyone can see whats going on, as opposed to process used by other orgs, prelim docs secret, only final result open
[05:56:20] <ggm> tech competence. we do stuff that works.
[05:56:25] <ggm> we do stuff that we have thought about.
[05:56:33] <ggm> take on work where we have people working, know what they are doing
[05:57:03] <ggm> organization made up of people as opposed to gov, orgs, other entities. core are the people who work here.
[05:57:24] <ggm> leadership chosen from those people, by those people, not by creating stds org that has org professionals, whose main goal is to grow the org
[05:57:43] <ggm> I think that the IETF realizes it cannot help the internet work better, would shut itself down.
[05:57:52] <ggm> and of course rough consensus and running code. we want stuff that works
[05:58:07] <ggm> those are the principles we adhere to, because they serve the community well. want to go on building on that basis
[05:58:24] <ggm> lots of words in prev. 2-3 slides, that people exchange all the time, use 5 times in one sentence.
[05:58:31] <ggm> one important part of document is definitions.
[05:58:57] <ggm> if we want to say 'its good for the internet' have to have definition of what the internet is. everyone has a def: the doc has one too.
[05:59:11] <ggm> when we talk about participants in the ietf we talk about people. not org, gov, but people who do engineering work
[05:59:48] <ggm> some of us, are not engineers. I respect them because they provide other strengths (?) but its a people thing. one of the twists which make the IETF work well. something which differs to many other orgs.
[06:00:12] <ggm> relevant means useful. was much longer, condensed to fit screen was odd.
[06:00:16] <ggm> must understand what we build.
[06:00:36] <ggm> lots of more text in the draft, very few who havent read yet. its important to the IETF to get this one right. feedback is very welcome.
[06:00:56] <ggm> this document doesnt have external dependencies, others depend on it. want to finish it soon. having a long discussion wont make it better
[06:01:02] <ggm> if possible, before the next IETF.,
[06:01:22] <ggm> Q to room. what is your opinion on this particular bit of restructuring hwo the IETF thinks of itself?
[06:01:31] <ggm> Stuart Bryant.
[06:02:22] <ggm> Have concern. intiial headline focus make internet work, will alienate people who use the suite to switch packets. may alienate some carriers. if this happens, if we don't address this concern, they will take their standards elsewhere. lot of evidence eg MPLS work in ITU.
[06:02:24] --- howard218 has left
[06:03:10] --- omarjan has left: Disconnected
[06:03:25] <ggm> Harald. have urgent request from SG15, following 27 docs want finished by november. there is a LOT more cooperation going on. also competition. but will ask you a Q. if the standards we make dont work on the internet, why should anyone believe they will work in a different venue?
[06:04:02] <ggm> Stuart. service providers work in a much more restrictire env. different set of requirements. relaxation of reqts working any to any on unregulateted system such as internet. one possibility is restriction. ITU work was moved because felt it would better address their interests, now no longer have design authority
[06:04:19] <ggm> Allison Mankin. Its johns WG.
[06:05:04] <ggm> we wound up sending pseudo wire drafts have been taken in by SG15. making them their documents too, making change, we haven't edited the change yet. no prior agreement, no editorial control, has to do with frustration. not clear how impacts, but sharing the turf issue. John is here.
[06:05:08] <ggm> John nothing to add to that.
[06:05:13] --- gih has joined
[06:05:29] <ggm> Harald yes there is an issue. have to put the internet first but doesnt mean we cant think about things. like to take that to the list.
[06:05:32] <ggm> James Polk.
[06:05:53] --- cabo-tzi-org has joined
[06:05:55] <ggm> Q posed to you leslie, didnt quite answer to my satisfaction. comments to docs. comments back to authors. Harald, I sent you a 3 pager, didnt even acknowledhe
[06:05:56] <hardie> Not so much "sending pseudo wire drafts" as "they took in pseudo wire draft" (note, that's my reading of what was said, and I have no knowledge of the situation)
[06:05:58] <ggm> Harald oops.
[06:06:10] <ggm> James. if you think I'm wet behind the ears, tell me.
[06:06:37] <ggm> host to host, interested me. things in groups I,m in have to *traverse* the internet. to some people thats ok. to some not.
[06:06:45] <leslie> James is James Polk.
[06:07:45] <ggm> comment that it doesnt work on the internet, whose to say it can work on a private domain. wrong. lots of thing within a single domain can be controlled internet may not want, or cannot do. pre-emption for instance. ITU got request to do Pre-emption last january. already standard. 3 years ago, I did a sip version, and its not even through the RFC ed. they did it in 12 months.
[06:08:01] <ggm> Harald. I see a mixture of Q.s one is I didn't reply to email on 10th feb. I will
[06:08:12] <ggm> James know you are busy. but if email list, could have ..
[06:08:24] --- ekr has joined
[06:08:43] <ggm> Harald. other issue. something which looks simple can take 4 years, thats a WG process failure. its either a reasonable thing to do, and should have been done sooner, or bad idea.
[06:08:53] <ggm> process failures are things we have to fix. thats my next talk.
[06:08:53] --- galvinjamesm has left: Disconnected
[06:08:54] <ggm> Ohta.
[06:09:30] --- omarjan has joined
[06:10:01] <ggm> afraid you think/say Internet first. I think you think IETF over Internet. I think you think neccessary. as I predicted more than 5 years ago IETF Yokohama. telephony over IP. just like that. Im afraid that Internet has first priority, IETF has second lower priority. I think if we want to move quickly and efficiently, IETF
[06:10:04] <ggm> will disappear.
[06:10:38] <ggm> Harald. I cant say I follow that. one thing I heard, we should put internet first, ietf second. I second that. If I was convinced the ITU wouild do better for the Internet Id move we close down.
[06:10:51] <ggm> Ohta. there are other possibilities, IEEE or orgs created by us.
[06:10:54] --- galvinjamesm has joined
[06:11:09] <anewton> brian rosen.
[06:11:32] <anewton> one word you could not apply to restructure is radical, evolutionary not revolutionary.
[06:11:35] --- Ole has left
[06:12:07] <anewton> people on IESG spend a lot of time on that job. NOMCOM says biggest problem is finding people with time
[06:12:28] <anewton> more radical change to stop IESG/IAB from spending so much time doing work.
[06:12:49] <anewton> harald.
[06:13:02] <anewton> I believe I am an engineer on temp. assignment.
[06:13:09] <anewton> fred baker.
[06:13:15] <anewton> I have returned to engineering.
[06:13:30] <anewton> IESG Chair is fulltime standards person.
[06:13:46] <anewton> We need people who can set aside my career for a period of time and then return to it.
[06:14:00] <anewton> When they do that, they need our support.
[06:14:11] <anewton> harald.
[06:14:32] <anewton> I want the process to work better so people can have a life, job, and IETF leadership... all three at once.
[06:15:04] --- Joe Touch has left: Replaced by new connection
[06:15:04] --- Joe Touch has joined
[06:15:05] --- Joe Touch has left
[06:15:12] <anewton> We need the engineers to be able to do this, has worked so far.
[06:15:30] <anewton> how many have read document.
[06:15:34] <anewton> room: some hands
[06:15:41] <anewton> harald.
[06:15:47] <anewton> who thinks going in wrong direction
[06:15:51] <anewton> room: one hand
[06:15:54] <anewton> harald.
[06:15:57] <anewton> right direction.
[06:16:01] <anewton> room: a few more hands
[06:16:08] <anewton> harald.
[06:16:09] --- howard218 has joined
[06:16:26] <anewton> if you have detailed questions on the doc, send them to Leslie and I. james, I will get back to you.
[06:17:01] <anewton> If need comments to overall group, send to IESG and IAB list. And there is always the IETF list.
[06:17:19] <hartmans> Are we not going to figure out how doomed we are financially today?
[06:17:25] <anewton> now acting as general area director
[06:17:53] <klensin-ietf> Observation: while this notion of permitting IESG folks to have lives is probably excessive, but we have already passed the point at which "engineers can do it" -- and also do anything else.
[06:18:02] --- Joe Touch has joined
[06:18:34] <anewton> this is about the core. about the working groups. about the documents. as randy bush use to say, it is documents all the way down.
[06:19:08] <anewton> an internet that works well, and a process that works well
[06:19:14] <anewton> process works slower than many would like
[06:19:48] <anewton> letting intellectuals stop the forward progress of technology, succumbing to rumor mongers with political agendas, etc. etc. etc.
[06:19:59] <anewton> have been accused to beholden to them
[06:20:14] <anewton> lots of problems, angry people. sign of something worng.
[06:20:52] --- Joe Touch has left
[06:21:00] <anewton> An IETF committed to improving.. keep an open consensus process. make bookkeeping just work. as fast as possible... but no faster!
[06:21:05] <mengwong> http://www.alvestrand.no/ietf/korea/Plenary%20Reform%20Status.pdf
[06:21:14] <anewton> but still want to think about docs
[06:21:25] --- zhangshushu has joined
[06:21:31] <anewton> ICAR
[06:21:40] --- avri has left: Disconnected
[06:21:49] <anewton> issues with "late surprize"
[06:21:54] --- rei has left: Disconnected
[06:22:03] <anewton> solution: review early, review often
[06:22:13] --- moc has joined
[06:22:35] <anewton> several experiments in review already exist.
[06:22:43] --- tuy has left
[06:23:16] <anewton> the working group is in start up mode.
[06:23:37] --- tuy has joined
[06:23:37] --- galvinjamesm has left: Disconnected
[06:23:49] <anewton> newtrk working group. standards track as defined doesn't match reality
[06:24:26] <anewton> important to get labeling right, because changing labeling can cause confusion
[06:24:28] --- ms has joined
[06:25:28] <anewton> PROTO. IESG review progress. too many documents to watch. too many working groups to watch. we need more hands - and common thoughts
[06:25:33] <anewton> wg chairs should be able to help
[06:25:40] <anewton> PROOT team tries to identify how
[06:25:43] <anewton> ongoing activity.
[06:26:45] --- Bob.Hinden has joined
[06:26:46] <anewton> WG process. difficulty in finding and sticking with consensus
[06:27:16] <anewton> difficulty in delaing with disruption. difficulty in tracking and resolving comments. difficulty in coordination outside WG
[06:28:12] <anewton> some tools are known to be useful. issue tracking. WG web pages. good minutes.
[06:28:28] <anewton> EDU team's WG efforts. http://edu.ietf.org/
[06:28:28] --- ft has left
[06:28:38] <anewton> some things are more difficult -- but try
[06:29:19] <anewton> but careful... there is a lot of baby in the bathwater.
[06:29:50] <anewton> proposals that make things better, but have limited impact on the global structure.
[06:30:11] <anewton> mailing list discipline. clarify rfc editor document handling in IESG
[06:30:21] <anewton> IESG review process. notification. documentation.
[06:31:11] <anewton> references Allison's presentation yesterday on statistics.
[06:31:11] --- avri has joined
[06:31:31] <anewton> PROTO role may change some roles of the IESG workflow.
[06:31:43] <anewton> building ICAR structures creates structures
[06:31:55] <anewton> the current proposals ....
[06:32:03] <anewton> this page intentionally left blank.
[06:32:27] <anewton> summary. we're doing the stuff that the IETF community now knows we need to do
[06:32:37] <anewton> we've started discussing some things we don't know how tod yet
[06:32:46] <anewton> we're trying to move quickly and be careful. this is hard.
[06:32:54] <anewton> your help is needed. it's HARD WORK!
[06:34:17] <anewton> questions, comments?
[06:34:22] <anewton> james polk.
[06:34:37] <anewton> concern with catching surprize late... what is "late"?
[06:34:39] <anewton> harald.
[06:35:01] <anewton> document wg last call and requires non-interoperable changes to spec
[06:35:05] <anewton> james.
[06:35:22] <anewton> I-D status is what proposed use to be years ago.
[06:36:04] <anewton> there are a lot of things sitting at proposed that people are implementing. people are confused by the standards track. newtrk is getting into this.
[06:36:07] <anewton> harald.
[06:36:20] <anewton> some people in newtrk say we should abandon the label RFC.
[06:36:21] <klensin-ietf> Except that draft-AnyDamnFool-position-00.txt is still an option for I-Ds and never was for Proposed
[06:36:33] <anewton> james.
[06:36:42] <anewton> concerned about high-bar for proposed standard.
[06:37:37] <anewton> harald.
[06:37:46] <anewton> spencer dawkins.
[06:38:08] <anewton> the idea wg will be able to do a working group snapshot.
[06:38:12] --- gih has left
[06:38:26] <anewton> it is an attempt to accomodate other things.
[06:38:43] <anewton> flexibiility given to the working group.
[06:38:46] <anewton> james.
[06:38:54] <anewton> status of entire working group.
[06:38:58] <anewton> spencer.
[06:39:06] <anewton> it is what the working group thinks it is
[06:39:15] <anewton> harald.
[06:39:23] <anewton> come to newtrk.
[06:39:25] <anewton> james.
[06:39:30] <anewton> beeen there, not getting much traction
[06:39:44] <anewton> needs to be area charter on each area.
[06:39:58] <anewton> what are the area's trying to accomplish.
[06:40:25] <anewton> iptel and mmusic is not obviously related to sip
[06:40:27] <anewton> harald.
[06:40:32] <anewton> finding stuff on the IETF website is hard
[06:40:36] <anewton> spencer.
[06:41:05] <anewton> refers to somebody wanting to implement 821 instead of 2821.
[06:41:15] <anewton> John L.
[06:41:48] <anewton> don't think IETF is always good at cost benefit analysis when doing standards.
[06:42:05] --- AWGY has joined
[06:42:07] --- tomphelan has left: Disconnected.
[06:42:09] <anewton> relates a spec that missed market window
[06:42:19] <anewton> need to consider this when doing reviews.
[06:42:26] <anewton> harald.
[06:42:28] --- mattz has left: Disconnected
[06:42:32] <anewton> need to do a lot more planning.
[06:42:52] <anewton> and keeping that consensus in front of people when they dive down there closest rathole.
[06:42:58] <anewton> ??
[06:42:59] --- mattz has joined
[06:43:13] --- AWGY has left
[06:43:24] <anewton> too much time needed to take something out there and take it back to historic.
[06:43:44] <anewton> suggests steps for things to automatically go historic.
[06:43:48] <anewton> harald.
[06:43:55] --- AWGY has joined
[06:44:20] <anewton> perhaps have someone else beside IESG do leg work.
[06:44:36] <anewton> ??.
[06:44:49] --- kdz has left
[06:44:53] --- memo56 has left
[06:45:14] <anewton> When something spans multiple areas, we tend to wait on doing the cross area review.
[06:46:10] <anewton> Training of wg chairs. The chair should guide the document all the way through the IESG process.
[06:46:18] <anewton> margaret w.
[06:46:35] <anewton> We have training for wg chairs but most don't come.
[06:46:56] <anewton> 1/5th or less of people are willing to spend the time to come.
[06:47:04] <anewton> joe touch.
[06:47:12] --- kurtis has left: Disconnected
[06:47:31] <anewton> If IESG is bogged down because of too many docs, stop watching them all.
[06:47:57] <anewton> Quit paying so much attention to Informational documents. Just check that it is not an end-run around wg's.
[06:48:00] <anewton> harald.
[06:48:09] <anewton> that is what was on my blank slide
[06:48:48] <anewton> (harald is digging in his laptop for a presentation)
[06:49:46] <anewton> IESG and the RFC Editor
[06:49:46] --- klensin-ietf has left: Disconnected
[06:50:11] <anewton> 10% of docs come to the IESG through RFC Editor.
[06:50:26] <anewton> IESG will no longer take responsibility of tech review for such docs.
[06:50:36] <anewton> Will still be responsible for end-run detection.
[06:50:53] --- omarjan has left: Disconnected
[06:51:25] <anewton> Must be careful to mark docs as such.
[06:51:42] <anewton> This change will be implemented soon!
[06:52:17] <anewton> Small Step suggested from John Klensin. Experimental procedure.
[06:52:29] <anewton> get from "think about it" to "try it" faster.
[06:52:36] --- omarjan has joined
[06:52:42] <anewton> even when this requires a change to our rulesets.
[06:52:57] <anewton> ohta.
[06:53:03] --- Joe Touch has joined
[06:53:48] <anewton> any one can build code for anything if they have enough money
[06:53:53] --- Joe Touch has left
[06:54:09] <anewton> how quickly will standardization process ???
[06:55:04] <anewton> Now saying that some RFC don't need IESG review. When? Need target.
[06:55:13] <anewton> harald.
[06:55:52] <anewton> two weeks for last call. telechats are every two weeks. so expecting from beginning to end in two months is setting expectations too high.
[06:56:15] <anewton> but unless problems are discovered, finishing in two months is good goal.
[06:56:18] <anewton> ohta.
[06:56:26] <anewton> how decides a problem is ??
[06:57:00] --- randy_g has left: Disconnected
[06:57:02] <anewton> fred baker.
[06:57:17] <anewton> 20 years ago there was not IETF.
[06:57:20] --- avri has left: Disconnected
[06:57:51] <anewton> IESG created to take load off of IAB
[06:58:22] <anewton> We are close to where we were at in 1992.
[06:58:46] <anewton> brian rosen.
[06:58:54] <anewton> I-D process could use work.
[06:58:59] --- ekr has left
[06:59:20] <anewton> Many of us are using a specific tool to make I-D's.
[06:59:47] <anewton> If you used an XML document to create a draft, we could almost automate the announcement/publication of it.
[07:00:18] <anewton> harald.
[07:00:21] <anewton> sounds like a good idea.
[07:00:57] <anewton> we are working with the secretariat on a different way of submitting I-D's. We should consider XML.
[07:00:58] --- tomphelan has joined
[07:01:03] <anewton> brian.
[07:01:15] <anewton> what documents the process and policy for submitting an rfc?
[07:01:46] <hartmans> Hmm, works in my wgs
[07:01:57] <anewton> Can I-D editor be made to announce to a specifi list.
[07:02:09] <anewton> What is the process for announcement?
[07:02:34] <anewton> bill f.
[07:02:57] --- BP has left
[07:03:02] <anewton> It is probably the case the some announcements are being swallowed by spam filters.
[07:03:10] <anewton> barbara.
[07:03:27] <jhutz> uh... 1ID-guidelines.txt
[07:03:46] --- narten has left
[07:03:48] <anewton> The issue of formats and how they are submitting will be working on when considering the new procedures.
[07:03:57] <anewton> Zafir.
[07:04:56] <anewton> Questions about drafts being bounced between working groups.
[07:05:16] --- mengwong has left: Disconnected
[07:05:21] <anewton> Sometime individual submissions are adopted by wg after it has been implemented.
[07:05:34] <anewton> harald.
[07:05:55] <anewton> Are you saying it is hard to call people's attention to individual submssions in I-D.
[07:06:01] <anewton> zafir.
[07:06:29] <anewton> Looking for improved process on what should happen to individual submissions.
[07:06:33] --- kivinen has left: Logged out
[07:06:33] --- kivinen has joined
[07:06:33] --- kivinen has left: Logged out
[07:06:46] <anewton> alex z.
[07:07:00] --- tuy has left
[07:07:07] --- klensin-ietf has joined
[07:07:13] --- moc has left
[07:07:24] <anewton> Issues with drafts going to more than one working group to make sure that it makes sense. I n some instances the working group cannot take on new work, not in charter.
[07:07:37] --- marushin has left: Replaced by new connection.
[07:07:38] --- marushin has joined
[07:07:38] --- marushin has left
[07:07:50] --- AWGY has left
[07:08:31] <anewton> Even though certain drafts don't make it on the charter, that does not mean it is not part of the IETF process. You can still work with members of a working group.
[07:09:13] <anewton> zafir.
[07:09:29] <anewton> Asking for more cooperation betwen wg's to get drafts accepted.
[07:09:34] <anewton> alex.
[07:09:53] <anewton> It is not just about what is in the pipeline. We should discuss off-line.
[07:09:55] <anewton> ??.
[07:10:04] <anewton> Move to a modern document format.
[07:10:07] <anewton> harald.
[07:10:18] <anewton> There is a mailing list dedicated to this topic.
[07:11:01] <anewton> alex.
[07:11:18] <anewton> I suggest we write a BCP on this to refer people to it.
[07:11:22] <anewton> Spencer.
[07:11:25] --- tomphelan has left: Disconnected.
[07:11:41] --- bkhabs has left
[07:12:17] --- kjd has left
[07:12:32] <anewton> Reference to John. K's draft.
[07:12:45] <leslie> draft-klensin-process-jul14
[07:12:49] <anewton> harald.
[07:13:10] <anewton> The mailing list on this is the solutions@alvestrand.no. please join.
[07:13:10] --- Bob.Hinden has left
[07:13:16] --- omarjan has left: Lost connection
[07:13:18] --- akihisau has left
[07:14:01] <anewton> We have a lot of work ahead of us, and sometime to know if we know if we have done it right, but it seems we are going in the right direction.
[07:14:08] <anewton> Thanks for coming.
[07:14:08] --- hartmans has left
[07:14:13] --- javier has left
[07:14:16] --- ggm has left
[07:14:17] --- hardie has left
[07:14:17] --- vlevigneron has left
[07:14:18] --- Loughney has left
[07:14:23] --- tskj has left
[07:14:25] --- howard218 has left
[07:14:27] --- sakai has left
[07:14:29] --- sob has left
[07:14:29] --- anewton has left
[07:14:39] <ms> <quit>
[07:14:49] --- jmc has left
[07:14:59] --- Bill has left
[07:14:59] --- leslie has left: Disconnected
[07:15:07] --- Lisa D has left
[07:15:13] --- cabo-tzi-org has left
[07:15:14] --- Glenn Parsons has left: Disconnected
[07:15:17] --- rlbob has left
[07:15:22] --- Ted Faber has left
[07:15:27] --- klensin-ietf has left
[07:15:50] --- sleinen has left: Disconnected
[07:15:53] --- ms has left
[07:15:54] --- ludomp has left
[07:16:07] --- warlord has left
[07:16:24] --- resnick has left: Disconnected
[07:16:44] --- tonyhansen has left
[07:16:49] --- lynn has left
[07:17:11] --- lha has left
[07:17:15] --- jhutz has left: Logged out
[07:17:15] --- jhutz has joined
[07:17:15] --- jhutz has left: Logged out
[07:20:31] --- arifumi has left
[07:27:16] --- MattMathis has left
[07:30:29] --- zhangshushu has left: Disconnected
[07:30:29] --- dcrocker has left: Disconnected
[07:33:49] --- shep has left
[07:34:36] --- mattz has left
[07:36:08] --- Netrangerrr has left: Disconnected
[07:59:14] --- LOGGING STARTED
[08:00:58] --- bonninjm has joined
[08:54:36] --- hta has joined
[08:55:35] --- hta has left
[09:28:57] --- bonninjm has left: Disconnected
[09:59:42] --- howard218 has joined
[10:08:00] --- howard218 has left
[14:10:21] --- randy_g has joined
[14:10:35] --- randy_g has left
[14:17:28] --- gih has joined
[14:17:36] --- gih has left
[17:35:29] --- resnick has joined
[17:35:49] --- resnick has left
[19:03:13] --- vlevigneron has joined
[19:03:29] --- vlevigneron has left
[23:29:03] --- sommerfeld has joined