[00:02:04] toshio.hiraga joins the room [00:02:19] Simon Perreault joins the room [00:02:42] Colman Ho joins the room [00:02:42] behcet.sarikaya joins the room [00:02:43] Lee Howard joins the room [00:02:54] shinmiyakawa joins the room [00:03:59] Atarashi Yoshifumi joins the room [00:06:10] Chris Donley joins the room [00:06:43] kawashimam joins the room [00:06:57] Chris Donley leaves the room [00:07:16] karen.s.seo joins the room [00:07:22] kawashimam leaves the room [00:07:43] sorry I was in the wrong chat room [00:07:48] agenda bash occurring [00:08:57] discussing whether or not to adopt ftp64 -- some folks say there was a consensus to do so and that the disagreement was on whether they like the document; others say I don't think there's consensus to adopt; group hum was split. [00:09:45] Thaler: group is chartered to handle translation. Yes, there may be a portion that's not on translation, but that would be something to discuss whether to remove it. (I'm not sure what the portion is) [00:09:54] Thaler speaking [00:10:12] Bo Zhou (china mobile) objecting [00:10:40] Thaler says let's not talk about what to do about the ALG text [00:10:40] Dave Thaler leaves the room [00:10:40] jlcJohn leaves the room [00:10:40] Colman Ho leaves the room [00:10:40] Lee Howard leaves the room [00:10:40] hirocomb leaves the room [00:10:40] shinmiyakawa leaves the room [00:10:40] Atarashi Yoshifumi leaves the room [00:10:40] Fuad Abinader leaves the room [00:10:55] Magnus Westerlund (Ericsson) [00:11:37] Magnus (AD?) agreeing with Thaler [00:11:50] Interim meetings [00:12:11] Thaler: have been having them monthly -- concentrated on first-priority scenarios -- nearly done [00:12:55] ylee joins the room [00:13:00] (Chair slides are available at https://datatracker.ietf.org/meeting/76/materials.html [https://datatracker.ietf.org/meeting/76/materials.html] [00:13:19] Thaler: Now going on to work on second set of scenarios [00:14:29] Upcoming WGLC -- list of docs -- want at least 5 positive reviews to progress documents -- looking for reviewers [00:14:57] Dan Wing (co-chair) -- looking for folks to review all the documents to check for inconsistencies [00:15:32] The point about cross-review is particularly important for those of us who know about DNS and not so much about the other issues, please! [00:15:47] Thaler emphasizing that it's *at least* 5 [00:16:02] Simon Perreault -- I reviewed all 5 and didn't find any problems [00:16:18] i *implemented* not reviewed [00:17:35] first talk -- IPv6 Addressing of IPv4/IPv6 translators -- I'm not sure who the speaker is X. Li? or C. Bao? [00:17:54] he's on slide 3 -- intro [00:18:23] satoru.matsushima joins the room [00:18:24] Last time agreed to include encapsulation [00:18:42] re: need for adequate terminology [00:19:14] becarpenter joins the room [00:19:32] "IPv4-translatable", "IPv4-converted" -- avoided "IPv4-mapped" and "IPv4-embedded" [00:20:06] slide 4 - Recommendations [00:20:14] Impact on Inter-Domain Routing [00:20:33] WKP (Well Known Prefix) [00:21:13] Referral support and optimal routing [00:21:33] Andrew Sullivan leaves the room [00:21:34] IPv6 address availability and consumption [00:21:40] Andrew Sullivan joins the room [00:21:55] Security [00:22:40] IPv6 address architecture [00:22:52] Support for multiple translators [00:23:06] slide 5 -- Address Format [00:23:57] Hajime Tazaki joins the room [00:23:57] Chris Donley joins the room [00:24:51] sureshk joins the room [00:24:54] slide 6 - Choice of Prefix for Stateless Translation Deployments [00:25:44] Speaker noted (on slide 5) that the U-bit creates a split in the address fields [00:26:24] slide 7 -- Choice of Suffix [00:28:17] Marcelo Bagnulo: You said "you must use all zeros" can't we use the null suffix for everthing else? [00:28:34] slide 7 -- Some open questions [00:28:53] sorry this is slide 8 [00:29:04] Marcelo: [00:29:45] "We're looking for input -- please express if there's anything specific you want when we add encapsulation? [00:30:35] Marcelo seems to be becoming the presenter [00:31:34] The only place where you can embed an IPv4 in an IPv6 (going back to slide 5 with format diagram) in the last 32 bits (as decimal) [00:31:44] He's pointing to the last 32 bits [00:33:03] we could also put the address -- gesturing elsewhere -- but this is for display [00:33:05] Thaler; [00:33:44] clarifying -- input already works; output will require changes to existing implementaions [00:34:36] Good to emphasize tools for debugging [00:34:42] Simon Perreault: [00:35:10] One requirement is to not change IPv6 code [00:35:12] Dan Wing: [00:35:33] We'll have to modify code no matter what [00:35:40] Brian Carpenter [00:36:43] liked the idea for a tool, but not elsewhere [00:36:44] Thaler: [00:38:30] Use of the bottom format would be useful for encapsulation and translation [00:38:38] Bagnulo again [00:38:55] Wing: [00:39:38] Another consideration as to where to put the IPv4 address (back on slide 4) -- will be use for security too [00:39:51] Bagnulo -- wll try to address folks' concerns [00:40:01] Wing: objects [00:40:19] slide 9 -- Choice of Prefix for Stateful Translation Deployments [00:41:54] (By the way -- I'm too new to this group to grasp what everyone is saying in real-time -- please feel free to jump in with clarifications and corrections to what I"m typing.) [00:42:17] karen.s.seo: you're doing a fine job! [00:42:21] slide 11 -- WKP: Reuse :: FFFF:0:0/96 [00:42:40] slide 12 -- WKP: Request a new /32 [00:43:25] jlcJohn joins the room [00:44:20] speaker observing that this may be only useful in some scenarios (bullet 3) [00:44:21] kawashimam joins the room [00:44:33] slide 13 -- WKP: Request a new /96 [00:45:27] slide 14 -- an intermediate approach [00:45:42] Fuad Abinader joins the room [00:46:24] hirocomb joins the room [00:46:26] Atarashi Yoshifumi joins the room [00:46:42] Erik Kline: [00:46:42] nnnshin joins the room [00:46:54] Could you combine these two [00:47:04] Wing objects [00:47:28] Thaler: [00:47:51] prefers to keep it simple, don't like the intermediate [00:48:06] between the other two -- leans toward /96 [00:48:19] NAT PT uses /96 [00:48:50] unless folks who've used it have had problems [00:48:57] Marcelo: [00:49:17] Thaler: [00:49:25] just talking about the default case [00:50:17] Hui Deng: [00:50:44] prefer shorter version [00:50:54] Marcelo -- you want to stuff things in the middle? [00:51:16] Hui raised another objection that I didn't catch [00:51:23] Marcelo -- 2 conflicting opinions [00:51:37] Alain Durand: keep it simple [00:51:43] Colman Ho joins the room [00:51:45] Alain -- /96 [00:51:58] Lee Howard joins the room [00:52:16] Thaler -- let's take hums [00:52:16] shamus joins the room [00:52:34] calvin joins the room [00:52:42] Chris Metz: keep it simple and not have multiple algs -- /32 or /96 [00:52:43] Vincent Levigneron joins the room [00:53:00] Hui Deng -- not sure what he said about /32 [00:53:14] Thaler : /32 vs /96 vs 2 algs [00:53:30] WKP should be /32 -- mild hum [00:53:34] hum [00:53:37] WKP should be /96 -- louder hum [00:53:48] WKP should be 2 algs -- no hums [00:54:34] rough consensus for /96 -- will still go to list [00:54:56] next talk "XLATE 03 updates and open issues" -- [00:55:33] skipping slides 2 and 3 as duplicated in previous talk [00:55:39] slide 4 -- Fragmentation handling [00:56:11] slide 5 -- Translator behaviou [00:56:22] Dave Thaler joins the room [00:56:22] "behavior" [00:56:55] Simon Perreault [00:57:23] translator is NOT a router -- it should just send MTU [00:57:26] Marcelo agrees [00:57:39] speaker again [00:57:51] Masashi joins the room [00:58:13] Alain Durand: [00:58:53] In Softwire -- got feedback from Transport folks that changing MSS was NOT acceptable. [00:59:21] speaker again -- should we not mention modifying MSS? [00:59:40] Wing: might be worth explaining why it's valuable -- all the vendors do. [00:59:45] tplunke joins the room [00:59:54] Durand: agree [01:00:17] Speaker again [01:00:29] Slide 6 Reduce fragments from IPv4 to IPv6 [01:00:58] slide 7 -- In the case of nonstandard implementations [01:01:12] tetsuya.innami leaves the room [01:02:12] Andrew Sullivan leaves the room [01:02:20] Andrew Sullivan joins the room [01:02:20] describing 2 problems -- can modify translator to help work around [01:02:24] Marcelo: [01:02:35] shenshuo joins the room [01:03:13] one situation is where admin has configured host to send packets only 1280 [01:03:15] norisuke_hirai joins the room [01:03:49] slide 8 -- Some IPv6 end system does not implement RFC 2460 [01:04:27] shinmiyakawa joins the room [01:04:37] shenshuo leaves the room: Replaced by new connection [01:05:03] Marcelo: [01:05:29] clarification -- ID was zero, in the proposal we copy the ID [01:05:42] Alain Durand [01:06:10] ylee leaves the room [01:06:15] you're trying to fix nodes that didn't read a spec (broken) why would they read this one? [01:06:18] Marcelo: [01:06:32] hosts are following the spec [01:07:25] but they're relying on having minimum of 1280 bytes MTU [01:07:33] Alain -- sort of agrees [01:07:35] Fred Templin: [01:08:28] (I'm not sure if Fred agreed or corrected the proposal) [01:08:34] Alain againg: [01:08:57] unhappy with breaking assumption for IPv6 that there is 1280 byte MTU [01:08:58] This sounds to me like we're trying to move from "translation" to "work around strictly correct, but still broken, site policy" [01:09:02] Remi Despres: [01:09:30] Wing: [01:09:56] FJB joins the room [01:10:26] Need to clarify the problem description -- it's NOT that the host received an ICMP too big it's that the network dropped the ICMP [01:10:33] Thaler: [01:10:39] Suresh Krishnan [01:11:01] Thaler [01:11:47] Should the translator take into account the fact that the host might not get the message? [01:11:55] (they're going back and forth) [01:12:17] Alain: [01:12:38] Wing: [01:12:53] Do we want to account ICMPv6 Packet Too Big? [01:13:14] Alain: It's wrong to send Packet too big for 1280 or less [01:13:17] Marcelo: [01:14:08] It seems to me to be enormous scope creep to move from "we're going to translate these packets" to "we're going to fix them up to deal with various strangers' firewall policies". I think the proposal opens a Pandora's box. [01:14:15] No, it's OK to send these ICMP too big messages and the spec says what the host [01:14:30] hum [01:14:44] stronger hum to NOT account for this [01:15:08] slide 9 again [01:15:34] Some IPv6 end system does not implement 2460 [01:16:10] Alain Durand: [01:17:05] Could this be analogous to a link which allows fragmentation/reassembly? [01:17:14] Thaler -- different affect on hop count [01:17:24] (I may have misunderstood Thaler's comment) [01:17:37] slide 10 -- ICMP/ICMPv6 extensions options [01:18:29] Alain was saying it's no different from allowing a router to fragment (at the link-layer) as long as it's reassembled onthe other side before it leaves the link. I said it's differnet because in that case it's 1 TTL hop (TTL only decremented once), but in the translator case each IPv4 hop shows up, so it's not "link-layer" [01:18:32] (Sorry -- Slide 9 was "Some routers fragment packets even if DF is set to one) [01:18:46] (Thank you, Dave) [01:19:00] Slide 11 -- Follow RFC 5509 (ICMP NAT) example [01:19:34] Lee Howard leaves the room [01:19:37] Lee Howard joins the room [01:19:41] Magnus Westerlund: [01:20:17] donley.chris joins the room [01:20:24] Wing: Any objections? [01:20:29] Chris Donley leaves the room [01:20:33] no objections [01:20:49] Slide 12 -- The differences from SIIT (RFC2765) -- speaker did not cover [01:21:01] Switching speakers to Marcelo [01:21:41] "Fragmentation issues in IPv4/IPv6 translation" [01:21:58] going to focus on stateful even though slides cover both stateful and stateless [01:22:14] CHENGANG joins the room [01:22:59] (I don't see these slides on the IETF web site) [01:23:21] Incoming IPv4 packet -- Full pkt, DF=0 [01:23:27] tetsuya.innami joins the room [01:23:46] Simon Perreault: [01:23:56] Can you make both situations valid? [01:24:08] Marcelo: [01:24:20] Andrew Sullivan leaves the room [01:24:27] Andrew Sullivan joins the room [01:24:49] Wel, if do both, we'd need to specify what state to keep, etc. [01:25:00] Fred Templin; [01:25:29] If host sent the packet with DF=0, let it be fragmented [01:25:34] Remis Despres: [01:26:40] Is host is supposed to receive PMTU too big? [01:26:57] Marcelo -- no, packet is supposed to be fragmented [01:27:13] next slide -- Incoming IPvr packet -- Fragment (I) [01:27:24] (I think Marcelo got the go ahead on previous slide) [01:27:36] donley.chris leaves the room [01:28:08] hold for first frag then send [01:28:09] (This is for IPv4) [01:28:21] Masashi leaves the room [01:28:31] next slide -- Incoming IPv4 packet -- Fragment (II) [01:28:51] Option 1 -- reassemble packet [01:29:17] simple but requires storage for storing fragments [01:29:46] next slide -- Incoming IPv4 packet -- Fragment (III) [01:30:19] Option 2 -- keep state on ID in the translator (from the first fragment) [01:30:31] Andrew Sullivan leaves the room [01:30:38] Andrew Sullivan joins the room [01:31:03] have to keep fragments if they arrive before the 1st fragment [01:31:21] if first fragment arrives before the others, store the requisite ID [01:31:49] pro: requires less memory, especially if first fragment arrive first or early [01:31:54] mawatari joins the room [01:32:14] FreeBSD does the second but doesn't store the out of order fragments [01:32:16] con: more complex, needs to deal with attacks (RFC3128, 1858) [01:32:35] Brian Carpenter: [01:32:46] I'm having trouble keeping up with where we are on the agenda. Is this draft-ietf-behave-v6v4-xlate-stateful-02? [01:32:59] yes [01:33:03] Do we need to decide? [01:33:10] Marcelo: [01:33:17] Carpenter: [01:33:34] (the main open issue with the xlate-stateful doc is just about fragmentation handling) [01:33:35] Couldn't you just specify the result, not the process? [01:33:50] (comment I typed was from Carpenter) [01:33:56] just present both options [01:34:11] Simon Perreault [01:34:12] the implementor can choose [01:34:30] Marcelo: [01:35:00] marka: is that for the mic? [01:35:09] (not that I can relay it :) ) [01:35:09] That was either (X Li or C. Bao) [01:35:16] Simon: so what should we do? [01:35:23] Marcelo -- drop it on the floor [01:35:27] Fred Templin [01:35:54] donley.chris joins the room [01:36:36] yes [01:37:01] ogud: Olafur Gudmundsson joins the room [01:37:08] Wing: [01:37:18] Suresh: [01:37:34] Hiroshi (?): [01:37:48] Hiroshi Miata at mic [01:37:58] FJB leaves the room [01:38:09] FJB joins the room [01:38:13] Marcelo: [01:38:13] Hiroshi Miyata at mic [01:38:37] Just present both options [01:38:43] Marcelo: how much guidance should we provide [01:39:16] I am sympathetic to Mark's point: present the options (outline all the considerations, for instance -- "here are all the attacks") and let the implementers choose [01:39:22] This is a bike shed [01:39:26] move on [01:39:34] it's not red enough! [01:40:12] Thaler: [01:40:46] please have discussion on the list [01:41:17] will also move to an interim (WEBX) [01:41:19] ylee joins the room [01:41:26] SImon Perreault [01:41:57] ꈲ joins the room [01:42:06] just do simple option for now [01:42:09] Magnus: [01:42:22] both are about as simple as each other [01:42:30] please clarify question [01:42:49] i said do the safe one, not the simple one [01:42:54] <ꈲ> I haven't participated in BEHAVE for a while — can anybody tell me why this has to be decided? Both seem great implementation options to me... [01:42:59] Marcelo: Same as question for stateless: [01:43:07] Fred Baker: [01:43:18] Yes, let's do the same thing as for stateless [01:43:34] norisuke_hirai leaves the room: Replaced by new connection [01:43:34] norisuke_hirai joins the room [01:43:54] Marcelo -- OK will replicate stateless solution -- any comments (none were heard) [01:44:43] (trying to bring "Andrew" on WEBX -- but we can't hear him) [01:45:41] sureshk leaves the room [01:46:09] next presentation -- on DNS 64 [01:46:16] Andrew Sullivan [01:46:34] Just one proposal -- sent to list [01:47:19] what should be done to address multiple interfaces (multi-homed) [01:47:47] and they're not all doing DNS 64 [01:48:31] 3 cases -- multiple IPv6 interfaces, dual stack (both to Internet), dual stack (Internet with v6, and internal with v4) [01:48:44] has sent proposed text to list [01:49:06] becarpenter leaves the room [01:49:07] Thaler: should proposed text be added to 02 [01:49:24] becarpenter joins the room [01:49:29] Thaler: everyone please post comments to the list [01:49:36] Marcelo: [01:49:51] Other thing that remains -- what about the address format. [01:50:12] I don't think this doc can move to WGLC until the other document is closer to done [01:50:36] Andrew: agrees that this doc refers to other docs that aren't ready; but this doc is done [01:50:37] Andrew Sullivan leaves the room [01:50:43] except for this issue [01:50:44] Andrew Sullivan joins the room [01:51:59] next talk: DNS46 for the IPv4/IPv6 Stateless Translator -- X. Li [01:52:22] slide 2 -- Objects of this draft [01:52:49] slide 3 -- Scenarios and stateless solution [01:53:57] slide 4 -- DNS function for the IPv4/IPv6 Stateless Translation [01:54:54] slide 5 -- DNS64 function for the IPv4/IPv6 Stateless Translation [01:55:04] Slide 6 [01:55:23] DNS46 function for the IPv4/Ipv6 Stateless Translator (1) [01:55:42] for those following the agenda, we're moving Dan's learn-prefix discussion to Friday [01:55:46] slide 7 -- DNS46 function for the IPv4/Ipv6 Stateless Translator (2) [01:55:49] mccap joins the room [01:56:17] slide 8 -- DNS46 function for the IPv4/Ipv6 Stateless Translator (3) [01:57:05] wmhaddad joins the room [01:57:07] slide 9 -- Remaining issues [01:57:40] Marcelo: [01:58:46] calvin leaves the room [01:58:48] hiroki.suenaga joins the room [01:58:53] 1st option should be done first, 3rd option last [01:58:56] becarpenter leaves the room [01:59:41] becarpenter joins the room [02:00:24] Hui Deng -- Scenarios and Requirements [02:00:35] slide 2 -- Generic Scenarios [02:00:43] 1. IPv6 only network [02:00:50] 2. Host to host communication [02:00:58] dapengliu joins the room [02:01:12] dapengliu leaves the room [02:01:23] 3. 3GPP TR 23.975 V0.1.0 (2009-09) IPv6 Migration Guidelines (release 10) [02:01:25] max joins the room [02:01:57] ogud: Olafur Gudmundsson leaves the room [02:01:58] 4. Open network platform [02:02:08] slide 3 -- scenario 1 (diagram) [02:02:51] (having slide display problems) [02:02:53] feature request for cisco: make webex understand ppt ascii art [02:03:51] feature request for microsoft: handle ppt better [02:04:03] ppt ascii art, that is [02:04:30] slide 4 -- scenario 2 [02:04:42] hiroki.suenaga leaves the room [02:04:48] It seems to me that ppt and friends noticed at some point that ascii art is very hard to use in order to draw diagrams [02:04:59] slide 3 -- Requirement background [02:05:08] Sorry that's slide 5 [02:05:12] and they drew the sensible conclusion that people would stop using it since there are thousands of better tools available [02:05:29] sureshk joins the room [02:05:37] slide 6 -- R1: Support for legacy IPv4 applications [02:05:48] (but maybe they ignored the IETF discussions on publication format!) [02:05:48] -- lots are not IPv6 aware [02:05:52] Andrew Sullivan: copy-pasting from an I-D is easier than recreating the diagram... [02:06:09] R2: Support IPv6 only networks [02:06:10] yes [02:06:23] R3 minimize overhead on wireless links (slide 8) [02:06:45] arifumi joins the room [02:06:51] where is the link for the presentation? [02:07:07] slide 9 -- R4: Allow decentralized host-to-host communication [02:07:19] Fuad Abinader leaves the room [02:07:33] ( I don't know, doesn't seem to be on the IETF web site for materials) [02:07:33] Colman: AFAIK it's not on the meeting site [02:07:39] I guess it wasn't uploaded in time [02:07:39] slide 10 -- Gap analysis [02:08:21] @colman: AFAIK, this presentation was not scheduled for today [02:08:29] that is why the slides are not there [02:08:31] shenshuo joins the room [02:08:46] row = each requirementDS-Lite, NAT64, GW-init-DS-Lite, DIVI (host), PNAT [02:09:01] sorry this is a table [02:09:21] rows are requirements, columns are the above listed items [02:09:27] Sri Gundavelli: [02:10:26] (I have no idea what the speaker said -- sorry) [02:10:33] Alain Durand: [02:10:49] No, it's Simon P [02:11:09] Sri said you need more columns to fully compare each approach [02:11:09] @karen: he believes that column 3 (GW-initiated DS-Lite should have Yes on all the rows [02:11:25] and more rows need to be added [02:12:31] (thank you) [02:13:06] Alain Durand (now speaking) [02:13:39] Andrew Sullivan leaves the room [02:13:46] Andrew Sullivan joins the room [02:14:15] calvin joins the room [02:14:34] (line is building up at the mic) [02:15:24] shenshuo is now known as Sean Shen [02:17:31] I think he's saying it would be better to solve the host-to-host problem using encapsulation (softwire) rather than translating/re-translating at both ends and "distorting" the packet. [02:18:06] that's what it sounded like to me too [02:18:12] (FWIW) [02:18:49] Li Xing: [02:19:00] seemed to be supporting something [02:19:03] Jari Arkko: [02:19:14] Woj.Dec joins the room [02:19:46] Li supported requirement R4 and said that DIVI actually supports (contrary to what's on the slide) [02:19:50] There are lots of scenarios -- perhaps should prioritize the scenarios and do the most important ones [02:20:41] maybe need a plan to do the simple ones first, more complicated later [02:22:36] issue with trying to run v6-only -- requires running 2 separate networks (one that supports PNAT, one for systems that don't) [02:22:40] Request to the chairs: Could the slides from *all* sessions (incl this one) be up loaded on time (ie before the session)? It's rather difficult to follow some of presentations without timely access to the slides [02:23:13] speaker is replying but I can't follow what he's saying [02:23:22] Tina Tsou (Ting Zou) [02:23:23] wmhaddad leaves the room: Computer went to sleep [02:23:29] shamus leaves the room [02:23:45] We implemented IPv4 in IPv6 -- we have running code [02:23:51] Basavarag Patil [02:24:02] Should add columns [02:24:22] (Sorry the name is Basavaraj) [02:24:28] Sri Gundavelli [02:25:02] I don't know who this is [02:25:14] Sri again [02:25:18] Andrew Sullivan leaves the room [02:25:27] Thaler [02:25:33] Andrew Sullivan joins the room [02:25:41] Is there interest in the scenarios and requirements? [02:26:46] ylee leaves the room [02:26:46] I haven't heard against the 4 requirements presented, or the scenarios. I heard comments re: adding requirements and adding columns and value in the cells in the GAP analysis [02:27:16] FJB leaves the room [02:28:01] I heard comments re: not all "yeses" are equal. General agreement that aiming for yes is good. Some argued for getting to yes via a particular path. [02:28:37] Charlie P. (Perkins?) [02:28:59] Agrees that there should be more columns [02:29:09] Simon Perreault leaves the room [02:29:14] sureshk leaves the room [02:29:16] toshio.hiraga leaves the room [02:29:27] behcet.sarikaya leaves the room [02:29:33] last slide "Next Step" [02:29:39] tetsuya.innami leaves the room [02:29:39] satoru.matsushima leaves the room [02:29:52] Sean Shen leaves the room [02:30:04] Question 1 -- Should we add the definition "host based translation" into the framework draft? [02:30:32] donley.chris leaves the room [02:30:34] Question 2 -- Should BEHAVE wg re-charter to cover the issues in host-based translation? [02:31:09] Marcelo: [02:31:24] Let's finish current documents [02:31:24] becarpenter leaves the room [02:31:37] Andrew Sullivan leaves the room [02:31:45] Andrew Sullivan joins the room [02:32:13] Let's identify the more critical scenarios [02:32:17] toshio.hiraga joins the room [02:32:38] Atarashi Yoshifumi leaves the room [02:32:39] Alain Durand [02:33:08] Agree with Marcelo -- stay focussed. Better to take this work to softwire and not duplicate their work [02:33:33] Lee Howard leaves the room [02:33:59] Li Xing -- agree with Marcelo -- finish current docs. I believe the current framework is compatible with implementation in home gateways [02:34:02] Sheng Jiang: [02:34:44] There are many "currents" from different ISPs. All the requirements/scenarios are important. I don't think we should stop working on them. [02:35:08] Gang Chen: [02:35:48] I'm not sure what he said but he seemed to want to be sure the host-based stuff was covered [02:36:04] Bo Zhou -- concerned re: closing the document [02:36:27] Marcelo -- clarify that we need to finish documents. [02:36:40] ꈲ leaves the room [02:36:50] Thaler -- closing document doesn't mean closing the discussion (he spoke prior to Marcelo, but I didn't type quickly enough) [02:36:52] norisuke_hirai leaves the room: Computer went to sleep [02:36:52] max leaves the room [02:37:05] Marcelo -- we can always write more docs [02:37:17] Speaker: [02:39:06] Tony Hain -- agree concept need to be closed; but if there are scenarios that aren't clearly documented, they should be documented and one shouldn't waste IESG time. Translator belongs here. Some of the solutions belong in SOFTWIRE. Leave the framework document here and clearly define the scenarios and NOT waste IESG time. Don't confuse scenarios with solution mechanisms [02:39:12] norisuke_hirai joins the room [02:39:13] Marcelo -- worried about further delays [02:40:12] kawashimam leaves the room [02:40:19] Ruri Hiromi leaves the room [02:40:21] Thaler -- will resume on Friday. [02:40:25] Bye [02:40:28] toshio.hiraga leaves the room [02:40:31] :) [02:40:33] nnnshin leaves the room [02:40:47] hirocomb leaves the room [02:40:56] CHENGANG leaves the room [02:41:06] tplunke leaves the room [02:41:41] Andrew Sullivan leaves the room [02:42:12] arifumi leaves the room [02:43:12] arifumi joins the room [02:43:13] Woj.Dec leaves the room [02:43:17] Woj.Dec joins the room [02:43:27] Woj.Dec leaves the room [02:43:57] Colman Ho leaves the room [02:44:05] karen.s.seo leaves the room [02:44:44] Hajime Tazaki leaves the room [02:46:24] marka leaves the room [02:46:40] calvin leaves the room [02:48:12] Fuad Abinader joins the room [02:51:51] Vincent Levigneron leaves the room [02:54:08] FJB joins the room [02:59:04] Dave Thaler leaves the room [02:59:27] shinmiyakawa leaves the room [02:59:31] norisuke_hirai leaves the room [03:10:33] arifumi leaves the room [03:12:09] max joins the room [03:23:21] ꈲ joins the room [03:45:28] ꈲ leaves the room [03:55:28] FJB leaves the room [04:01:40] arifumi joins the room [04:02:00] mawatari leaves the room [04:02:29] arifumi leaves the room [04:06:05] Ruri Hiromi joins the room [04:06:33] norisuke_hirai joins the room [04:06:48] mccap leaves the room [04:10:52] FJB joins the room [04:12:19] norisuke_hirai leaves the room [04:13:41] Vincent Levigneron joins the room [04:14:37] FJB leaves the room [04:16:19] Woj.Dec joins the room [04:19:38] FJB joins the room [04:21:47] Woj.Dec leaves the room: Replaced by new connection [04:21:48] Woj.Dec joins the room [04:26:18] max leaves the room [04:42:45] Woj.Dec leaves the room [05:01:24] FJB leaves the room [05:01:28] FJB joins the room [05:07:31] FJB leaves the room [05:07:43] FJB joins the room [05:08:32] FJB leaves the room [06:00:51] Ruri Hiromi leaves the room [06:02:33] Fuad Abinader leaves the room [06:14:30] Ruri Hiromi joins the room [06:22:59] Vincent Levigneron leaves the room [07:11:17] Ruri Hiromi leaves the room [07:33:42] Ruri Hiromi joins the room [07:58:24] wmhaddad joins the room [08:03:34] wmhaddad leaves the room [08:24:58] jlcJohn leaves the room [08:56:00] Vincent Levigneron joins the room [10:49:39] Ruri Hiromi leaves the room [11:09:39] Vincent Levigneron leaves the room [15:12:21] Ruri Hiromi joins the room [15:12:21] bnsmith leaves the room [22:43:33] Ruri Hiromi leaves the room