[00:03:33] dougm.home joins the room [00:46:37] dougm.home leaves the room [06:33:20] martin.thomson joins the room [06:38:44] yone joins the room [07:01:12] Joe Hildebrand joins the room [07:02:37] g.e.montenegro joins the room [07:03:08] Barry Leiba joins the room [07:03:55] fujiwara joins the room [07:04:28] tonyhansen joins the room [07:04:44] tonyhansen has set the subject to: AppArea General Meeting, IETF 78 [07:05:01] Reminder: please say your name if you come to the mic. [07:05:13] healthyao2000 joins the room [07:05:26] NomCom: nominations for Apps AD [07:06:30] coopdanger joins the room [07:06:49] Agenda bashing. [07:08:13] stpeter joins the room [07:08:28] mtcarrasco joins the room [07:08:58] Demo of the IETF datatracker [07:09:09] @joe: no bashing I hope? [07:09:31] martin.thomson: none [07:09:44] ta [07:10:08] sm joins the room [07:10:47] Mark Nottingham joins the room [07:12:37] From the floor, JohnK says "try showing the IDNA drafts" [07:13:37] Back-and-forth about the actual comments, no net semantic content. [07:14:03] Atarashi Yoshifumi joins the room [07:14:49] Yves Lafon joins the room [07:16:12] Murray: there used to be a page that showed all of the state transitions. Has that disappeared? [07:16:48] Alexey: I'll follow up [07:17:10] Barry Leiba: work going on to expand the datatracker for non-WG docs. [07:18:24] BoF announcemnents [07:18:45] Jeff Hodges to talk about HASMAT [07:19:18] Randall Gellens joins the room [07:19:36] Tomorrow at 1pm [07:19:55] HTTP Application Security Minus Application and Transport [07:20:03] !authenticaiton [07:20:10] !tls [07:20:52] Web 2.0 organic growth, intense complexity, gnarly interactions between HTML[5] and javascript, redirects, etc. [07:21:16] Lots of attacks, such as XSS (etc) [07:21:28] ob joins the room [07:21:56] no coordination between the different security responses at different layers. [07:22:11] EKR joins the room [07:22:22] 3 I-Ds: Same Origin, strict transport security, media type sniffing [07:22:31] resnick joins the room [07:22:34] requirements doc [07:22:43] coordinate with w3c. [07:22:54] W3C WebApp Security WG [07:23:22] (someone please post in a link to Jeff's last message to apps mailing list) [07:24:13] Alfred Hoenes at the mic [07:24:16] Alfred Hoenes about URNs [07:24:35] eburger joins the room [07:24:49] most important use is in libraries [07:25:20] URNs are only informational/experimental [07:25:34] (aside: really? i guess this *is* important) [07:25:56] in particular, ISBN13 [07:26:20] National Bibliographic Names (NBN) needs updating [07:27:04] RFC 2141 not up to date wrt ABNF, etc. [07:27:50] i think i misunderstood above about info/experimental; i think he was talking about specific URN schemes such as ISBN. [07:28:38] they're open to taking on other schemes, but the original charter has only library-related ones. [07:28:52] narten joins the room [07:29:30] Joe Hildebrand: new URN schemes, too? [07:29:38] Alexey: out of scope, question for the BoF. [07:29:50] Barry Leiba: FTPEXT2 [07:30:32] ~6 drafts as extensions to FTP [07:31:02] Today at 17:40 in Colorado [07:31:42] Patrik Faltstrom: Bar BoF (breakfast) on service discovery [07:32:31] URI DNS RR I-D for example [07:34:05] Examples: A/AAAA, MX, SRV, DDDS, etc. [07:34:13] Wednesday 8AM [07:34:48] Interest from the CORE wg [07:35:07] paf will check to see if we can use 2.1, if possible. [07:35:22] will post to apps-discuss when we know for sure. [07:35:50] Errata processing for apps-area RFCs [07:38:10] Expect emails from the ADs to various WG mailers asking for help [07:38:26] Tony: why 2821 is on the list? [07:38:54] stpeter: people have been filing errata for years. [07:39:36] stpeter: the errata may be moot at this point, but need to be dispositioned. [07:40:06] Barry: there should be a process for ensuring that when a doc is obsoleted, the errata need to be addressed. [07:40:29] for example, nobody was notified that there were errata on 2821 when it was obsoleted. [07:40:49] JohnK: the management item should be at last call time [07:41:26] JohnK: RFC editor is accepting errata on obsolete docs. that shouldn't happen [07:41:44] eliot.lear joins the room [07:41:53] 2821 errata should have been cleared, since they were checked. [07:41:57] sgerdes joins the room [07:42:12] stpeter: many process improvements needed [07:42:28] i was thinking of high voltage electrodes to close items ;-) [07:42:31] tooling, process for bis drafts, etc. [07:43:08] (note: 3920bis/3921bis need errata checking) [07:43:32] Randall: there are other changes requested that are not errata. Feature requests, e.g. [07:44:09] Alexey: "Hold for update" state for errata might work [07:44:49] Dave Crocker: "errata" = "request for comments" [07:45:01] i.e. these are the comments that were requested [07:45:14] Julian joins the room [07:45:18] DaveC: what about my draft? [07:45:33] (missed his RFC number) [07:46:25] ray_atarashi joins the room [07:46:38] Cyrus: i want a wiki page for every RFC [07:46:52] Julian leaves the room: Computer went to sleep [07:46:52] (wow. i like that idea) [07:47:01] Alexey: tools.ietf.org has issue tracker [07:47:16] Cyrus: needs to be more visible than that [07:47:32] Alexey: send reminder, or message to tools-discuss [07:48:30] From the back of the room: should there be links to errata from the tools version of the draft? [07:48:37] Mark Nottingham leaves the room [07:50:04] Olafur: announcement of DNSext work on i18n [07:50:15] (Olafur opens a can of worms) [07:50:42] IDNA2008 implementation [07:50:43] Dave Cridland joins the room [07:50:51] Yoshiro YONEYA [07:51:10] Spec is fixed, RFC numbers pending [07:51:35] RFC 5890-5893 [07:51:43] Assigned, but not published yet [07:51:51] 2 mapping mechanisms [07:52:00] draft-resman-idna2008-mappings-01 [07:52:08] Unicode TR 46 [07:52:21] does not break punycode labels [07:52:26] EKR leaves the room [07:52:27] idnkit-2.0 [07:52:51] input -> mapping(optional) -> idns2008 -> output [07:53:39] (Note to folks implementing these text transformations: check to see if "ropes" help) [07:54:12] API for apps: U->A, A->U, compare, valid? [07:54:18] C only, Java/Perl/Python coming [07:54:28] (Joe Hildebrand, You mean the string-like class commonly lurking in some C++ libraries, or something else?) [07:54:57] (Dave, yes, the string-like thing that uses trees) [07:54:59] Mark Nottingham joins the room [07:55:14] http://jprs.co.jp/idn/index-e.html [07:55:27] (Dave: i was just reading about them, which is why they were top-of-mind) [07:55:52] (Joe Hildebrand, Your current hammer? :-P) [07:56:01] (Dave: not yet. :) ) [07:56:28] tonyhansen leaves the room [07:56:41] paf: apologize, have idna2008 impl, but ugly [07:56:59] normalization of korean characters is using cheats [07:57:31] (someone please repeat the speaker's name here) [07:57:46] HTTP authentication is week in security and functionality [07:57:58] client crts too complex. [07:58:18] no log-off, modal dialog, no guest users [07:58:38] Resnick: users? or implementers? [07:58:48] Speaker: users [07:58:49] (Very confusing, the lag must be about two or three minutes on the audio. I wonder if how much of this is local.) [07:58:55] ob leaves the room [07:58:59] key management, public key complexity [07:59:01] obergmann2 joins the room [07:59:16] barry: different client certs for different computers, different for different apps [07:59:28] resnick: modal dialog is a problem? [07:59:38] speaker: spec doesn't require, but that's how it's implemented [07:59:57] speaker: we can't control the bad implementations [08:00:30] plain text only, weak against phishing [08:01:01] proposal: mutual authentication for HTTP [08:01:08] Mark Nottingham leaves the room [08:01:16] secure <-> http basic/digest [08:01:23] easy to use [08:01:31] mutual [08:01:55] implemented on top of RFC 2617 [08:02:02] PAKE as underlying crypto [08:02:10] authentication only [08:02:28] client side: no keys/storage required [08:02:39] server side: user/secret table required [08:02:47] drop-in replacement for basic/digest [08:02:57] asserts that it's easy to deploy [08:03:08] Am I the only one wondering why HTTP requires its very own authentication suite? [08:03:24] (dwd: no. the scribe is wondering the same thing) [08:03:35] narten leaves the room [08:04:03] Someone fancy asking why we don't put effort into using SASL or GSSAPI in HTTP instead? [08:04:04] Trusted display for mutual auth. firefox plugin [08:04:18] (ekr is hovering near the mic. wait for it) [08:04:31] draft-oiwa-http-mutualauth-06 [08:04:55] impl: apache module, webrick/ruby, mozilla patch, ruby client [08:05:12] Yves Lafon leaves the room [08:05:14] sm leaves the room [08:05:15] g.e.montenegro leaves the room [08:05:40] (where is the venue for discussion?) [08:06:19] missed his name, from the back: you still use passwords? [08:06:31] Use SSO instead? [08:06:59] speaker: only encrypted passwords go to server [08:07:29] ekr: password-based security in browsers is ghastly. agree [08:07:40] ekr: other existing mechanisms also bad. [08:07:47] (That's not EKR, it's far too slow...) [08:08:09] ekr: why would sites use this? they won't be able to control the UI [08:08:17] Cary joins the room [08:08:37] speaker: phishing. [08:08:56] healthyao2000 leaves the room [08:09:15] sm joins the room [08:09:25] g.e.montenegro joins the room [08:09:26] healthyao2000 joins the room [08:09:34] ekr: there are a number of technical authN mechanisms. DIGEST, TLS-SRP, etc. [08:09:56] not in use because they invade the user experience [08:10:06] Alexey cuts discussion [08:10:16] sgerdes leaves the room [08:10:20] eliot lear: related issue for non-web authentication [08:10:20] Yves Lafon joins the room [08:10:39] FedAuth BOF tomorrow morning [08:11:17] Alexey asks folks to be nice to the newbie. [08:11:35] Ulrich Konig: privacy prefs for e-mail messages [08:11:47] German gov agency for enforcing privcay laws [08:11:56] coopdanger leaves the room [08:12:09] EKR joins the room [08:12:36] how does sender influence what receiver does with your email [08:12:48] disclaimer, prose, digital restrictions [08:12:59] barry, clarifying: you mean like "please don't print/forward" [08:13:01] yes [08:13:25] suggestion: Privicon [08:13:37] [X] - keep secret [08:13:44] [=] delete after reading [08:13:45] (e.g) [08:14:00] parse: first line, subject, header [08:14:05] Julian joins the room [08:14:31] So this is a globally defined security policy with FLOT labelling and a universal clearance? [08:14:35] sgerdes joins the room [08:14:45] bear code [08:14:50] (and implicit trust in the receiver) [08:14:55] need "burn after reading" [08:15:05] Igo: you expect the client software to enforce this? [08:15:10] burn BEFORE reading [08:15:19] ;-) [08:15:21] speaker: no. plugin can help, but otherwise it's plain text [08:15:29] Barry Leiba, Burn before decoding this weird code? [08:15:29] (barry: lqtm) [08:15:56] [>] please share [08:15:56] this seems like a great idea [08:16:11] (please retweet) [08:16:13] eliot.lear, Which is why we actually have specs for this. [08:16:27] what are they? [08:16:30] Be nice :-) [08:16:34] eliot.lear, ESS, etc. [08:16:40] ref, please? [08:16:52] Randall: suggestion, don't talk about icons, but more about functionality. who is interested and why? [08:17:45] eliot.lear, Extended Security Services for S/MIME. This basically defined your labelling format. This particular incarnation is a new label format, and uses FLOT labelling (first line of text). [08:18:02] eliot: as a user i like the idea [08:18:13] (the i there was referring to eliot) [08:18:29] John C Klensin joins the room [08:18:44] wallstreet: also notes to MTAs? [08:19:14] paf: interesting, but can't even do reply marking [08:19:21] so i would require S/MIME to do this? [08:19:32] barry: p3p to define policies for websites [08:19:49] speaker: overlap in interest with p3p [08:19:55] eliot.lear, Yes, although in practise, FLOT is more common, which doesn't, and just inserts the displayMarking of the label into the first line of the message. [08:20:01] Kev joins the room [08:20:09] JohnK: 40y experience in encoding things in subtle ways [08:20:38] JohnK: only 0 have worked. (with ironic tone) [08:20:48] JohnK: except for almost re: [08:20:59] eliot.lear, In my evil work life, we see this quite a bit. In XMPP, we actually do it properly. The advantage of using a displayMarking as a fallback is that users see "UNCLASSIFIED" and know (mostly) what it means. [08:21:41] name? [08:21:57] That was Murray Kucherawy [08:22:00] name: apps wg to work? [08:22:23] Dave Crocker: interesting. existing proof w/tiny community. [08:22:36] DaveC: conventions into email, encoding (human factors) [08:22:56] DaveC: Murray's mother couldn't understand [08:23:11] it was in the dix bof, and it was my dad [08:23:18] and it was a lie. it was me [08:23:21] (lots of potential yo-mamma jokes float in the air, but nobody says out loud, just barely) [08:23:40] EKR leaves the room [08:23:47] EKR joins the room [08:23:48] sampo: will regulatory agency come after me if i don't follow? [08:23:57] speaker: no. just a social convention [08:24:25] Hannes: how did previous tries fail? p3p for example [08:24:44] alexey guards the mic. [08:25:24] Javier Ubillos: name-based sockets [08:26:19] martin.thomson leaves the room [08:27:05] two typical approaches: HIP/socket abstractions [08:27:16] Surrogate addresses: new namespace [08:27:32] OS gives multi-homing, etc. [08:27:44] Apps can't tell if this is happening, so they have to re-implement. [08:27:59] Socket abstractions [08:28:08] one for every framework [08:28:17] resolve once, reuse IP [08:28:25] What do we want? [08:28:42] no new indirections, no new delays, address management, backwards-compatability [08:28:55] coopdanger joins the room [08:29:17] coopdanger leaves the room [08:29:19] Addr mgmt (mobility, multi-homing, renumbering, v4/v6, NAT penetration) [08:29:27] HIP adds delays, for example [08:29:51] API [08:30:02] fd = socket (AF_NAME, SOCK_STREAM, 0) [08:30:40] accept(peer_name, fd) [08:30:47] Initial name exchange [08:31:25] oh… this is somewhat like the session reestablishment from XEP-198, but down a layer. [08:31:41] Joe Hildebrand, The XSF is always ahead, you see... [08:31:51] martin.thomson joins the room [08:32:02] eburger leaves the room [08:32:33] impls for ubuntu/android [08:32:47] Support for Shim6 [08:33:04] Working on support for UDP [08:33:36] invite to internet area where we'll talk more about shim6 [08:33:59] roadmap: [08:34:23] v4/v6 interop, NAT penetration, path diversity utilization, naming resolution (host/application, etc) [08:34:41] pete resnick: relationship between this and multipath TCP? [08:34:47] JLCjohn joins the room [08:34:56] pete resnick: also compare w/Apple's API for this [08:35:10] speaker: API options vs. path diversity [08:35:30] tonyhansen joins the room [08:35:35] potential that this is the API for MTCP [08:35:51] re Apple: takes action item [08:36:02] eliot.lear leaves the room [08:36:14] JohnK re: deprecating Unicode language tag chars [08:36:55] draft-presuhn-rfc2482-hostoric-02 [08:37:07] unicode has deprecated already [08:37:52] Mark Nottingham joins the room [08:38:44] barry: i support this [08:38:55] martin.thomson leaves the room [08:39:26] martin thomson: http timeouts [08:40:13] request timeout. long-polling. how long to hold a request open [08:40:29] servers either have to guess, or application has to provide [08:40:43] barry: don't you hold these open for hours? [08:40:50] paf joins the room [08:40:56] martin: intermediaries drop connections [08:41:11] (note: sometimes without a FIN or error) [08:41:48] want the timeout to be long, but finding out what works is hard [08:41:55] timeout* header [08:42:08] Request-Timeout header [08:42:27] intermediaries can reduce according to their knowledge/policy [08:42:41] origina server sees lowest value [08:42:44] origin [08:43:58] Second part: dealing with idl connections [08:44:40] http/1.1 connections are reusable in theory [08:44:45] in practice not so much [08:44:54] tlyu joins the room [08:45:01] many clients always make a new connection for a POST [08:45:14] if you want to reuse, how can you tell if the connection is still valid? [08:45:26] Both ends can notify about policy [08:45:41] hop-by-hop header [08:45:52] both peers advertise how long [08:46:00] Connection-Timeout header [08:46:10] Keep-Alive header suggestion [08:46:31] existing implementations are inconsisten [08:46:33] t [08:46:59] apache sets n/v pairs [08:47:06] firefox uses "115" [08:47:24] draft-loreto-http-timeout-00 [08:48:02] alexey (from private question) why not in HTTPbis? [08:48:09] not on charter, wide interest [08:48:24] How much of a fork-lift upgrade is this, and if this could be deployed effectively, would the original problem go away? [08:48:29] apps-discuss for now [08:48:45] Mark Nottingham leaves the room [08:48:46] Dave: you want that in-room? [08:48:51] Joe Hildebrand, Please. [08:49:06] Mark Nottingham joins the room [08:49:09] from floor: TCP user timeout? [08:49:19] Maybe we should just say HTTP 2.0 is TCP. [08:49:44] (note: TCP stuff is hop-by-hop, not end-to-end, so doesn't solve the problem) [08:49:44] Randall Gellens, Can't, not nearly verbose and broken enough. :-) [08:50:05] cyrus joins the room [08:50:15] martin: client and server aren't problem, it's the intermediaries. [08:50:24] Yeah, plus it doesn't magically go through firewalls, nor can it be done in Javascript. [08:50:53] martin: not looking for forklift [08:51:05] Next agenda item [08:51:10] Apps Area WG proposal [08:52:28] Consensus calls harder w/o WG [08:52:58] martin.thomson joins the room [08:53:09] I am now very very late with a comment on the "email tagging issues". My comment at the microphone about "reply" issues is that some clients do translate the "Re: " tag to (for example) "Sv: " if you have the client/environment in Swedish Locale. When brought up with vendor, they claim that is what the RFC is saying they can/should do. My reading says "only Re: can be added, if anything". Vendor say "unless RFC exists, we will not even think of changing how the client works". [08:53:52] (paf: agree. it's particularly apparent now that I regularly converse with folks in China) [08:53:56] So email threads in Sweden often has "Re: Sv: Re: Sv: Re: Sv: the subject" as Subject. [08:54:27] martin.thomson, What I meant, BTW, is that if you fix the intermediaries to handle timeout headers, would this also fix them to not cause the major problems they are with dropped (and severed) TCP connections? [08:54:33] AW: indeed [08:54:49] Re: subject line is freeform [08:54:56] various elder-statesmen start to wander towards mic. [08:54:58] dave cridland: I'm sorry, I misunderstood your question :) [08:55:18] martin.thomson, Or avoided it with the deft touch of a born statesman. ;-) [08:55:32] Indeed, "Re:" (and to a lessor extend, "Fwd") have been very problematic [08:55:46] davec: probably not - you need the ability to drop a connection when it's idle, I don't think that you can avoid that [08:55:58] you also need to be able to drop a connection to shed load [08:56:04] IMAP THREAD and SORT have algorithms for extracting a "canonical" subject for sorting/threading purposes. [08:56:08] so I don't see a way to avoid the general problem [08:56:11] The intent is that the exact string "Re:" is used to indicate reply, but mixing protocol and user information is always a problem [08:56:14] (martin: MUST send FIN at least) [08:56:34] Also there is the whole business of mailing list "[XXXX]" prefixes. [08:56:35] BarryL: used to push this, until we had apps-discuss. still supports [08:56:48] martin.thomson, SO you're saying that intermediaries will find the timeout stuff useful, and this usefullness will benefit all the players even after they're updated. That's good. [08:56:50] @joe: FIN is a nicety that you don't always get, but it would be nice to see that instead of RST [08:56:57] BarryL: charter should say they can spin off other WGs as needed [08:57:17] (martin: sometimes you don't get EITHER) [08:57:34] daveC that's the idea, can I employ you for marketing, that was great [08:57:39] @joe tell me about it [08:57:47] BarryL: wouldn't need to meet separately, could use this slot [08:58:15] martin.thomson, Okay, I accused you of being a politician, but marketing? That's low. [08:58:32] Barry: sometimes we might not need a BoF if it came out of the Apps Area WG. [08:58:51] DaveC: :D [08:58:53] Resnick (as Don Quixote) [08:59:15] Resnick: offloading work from ADs is a good thing [09:00:13] Resnick: except. we should have been doing this for 20 years. you're saying that WGs are too heavy weight. [09:00:36] Resnick: why not spin up small WGs? because we don't have mechanism to deal with this at IESG. [09:00:58] resnick, You're saying this is a workaround hack, and not fixing the problem that's at the core. [09:01:02] Resnick: couldn't we just fix the process? [09:01:12] (dave: yes) [09:01:19] sm: Re: is special. see 3.6.5 of RFC 5322: 'When used in a reply, the field body MAY start with the string "Re: " (an abbreviation of the Latin "in re", meaning "in the matter of") followed by the contents of the "Subject:" field body of the original message.' [09:01:26] Dave Crocker: ack the discomfort [09:01:37] resnick, We should form a BOF to discuss this idea, perhaps with a view to considering forming a working group to discuss this adaptation of the theoretical policy. [09:02:23] DaveC: doesn't fix the underlying problem [09:02:24] has anyone talked to Robert and Gonzalo (RAI ADs) about their success with DISPATCH? [09:02:45] DaveC: apps is too diverse, consensus will be hard [09:02:59] Joe Hildebrand, The Daves need at least three characters from the surname for distinguishing purposes, sorry. :-) [09:03:14] crocker [09:03:18] sorry [09:03:27] Apropos to Quixote and being in the Netherlands: http://xkcd.com/556/ [09:03:35] dcrocker: used to be possible to have individual submissions [09:03:50] marc.blanchet.qc joins the room [09:03:58] dcrocker: the IESG is the barrier to all of the lightweight processes. [09:04:35] Haibin joins the room [09:04:38] murray: individual drafts hard to generate enough consensus to get through IESG [09:04:57] some of this discussion might be better in a plenary [09:04:58] murray: does that mean no more individual drafts? [09:05:21] alexey: hard to tell if any individual draft is going to be a problem child [09:05:45] stpeter: what about a mailing list per document? [09:06:04] No mailinglist ... wikis [09:06:14] cyrus: formal app-area consensus call for each doc? [09:06:25] stpeter: that's what the apps wg is [09:06:36] FWIW, I'm happy to effectively kill individual submissions as long as firing up a working group (or at least a mailing list for discussion) is sufficiently lightweight. [09:06:55] (missed a couple of those last exchanges w/murray) [09:07:56] Barry: dcrocker said we shouldn't need working groups to achieve consensus. barry wonders where we're supposed to discuss, so we can show the IESG that consensus was reached [09:08:49] dcrocker: in favor of online discussion, tracking for consensus [09:09:03] dcrocker: caught in a timewarp of 20y ago [09:09:21] dcrocker: consensus now means 1 person can veto, didn't used to be [09:09:59] Mark Nottingham leaves the room [09:10:49] dcrocker: wg process allows labeling something as consensus, but doesn't change the actual amount of consensus [09:11:20] EKR leaves the room [09:12:13] leslie: the IESG is not the only reason that things are slow [09:13:54] lesle: proposal: start this as a 4week experiement [09:14:48] obergmann2 leaves the room [09:16:10] (wow. i just lost attention span. HELP) [09:16:29] EKR joins the room [09:16:56] speaker: should use different procedure for this wg [09:17:41] speaker is Jiankang Yao [09:18:08] I think he was confusing individual submissions with independent submissions. [09:18:13] tony hansen: This kind of wg could help as a clearing house for where to redirect individual submissions. [09:18:16] martin.thomson leaves the room [09:18:21] (pete: thanks!) [09:18:30] alexey: More like RAI dispatch. [09:18:38] alexey: This is a different model. [09:18:57] dave crocker: If someone doesn't know where to have the discussion, how do they know to come here. [09:19:18] dave crocker: basic problem is that folks have to garner support. Doesn't matter where that is done. [09:19:37] dave crocker: Nothing proposed changes that requirement. It's just about where to do it. [09:19:55] dave crocker: This proposal might diffuse efforts at support. [09:20:43] dave crocker: this is not a place for focused work. (e.g., for mail work, putting it on ietf-smtp is a good thing. apps-area is wrong place.) [09:20:51] dave crocker: not seeing the value-add. [09:21:19] dave crocker: experiment not really an experiment. [09:21:36] I suspect that one thing that may well help is if all individual submissions were announced on the apps-discuss list early on, to draw (and highlight) attention. [09:22:18] leslie daigle: proposal doesn't stop individuals from doing work outside of this wg. This wg is for a review function at the end of that indivdual work. [09:22:59] and also to point out the discussion mailing list address in the boilerplate [09:23:01] Good question... [09:23:16] Ideally people would figure it out earlier on than that. [09:23:31] resnick suggested that I was being Don Quixote for a reason: we should probably just try this because the alternative is fixing something that has been broken for years. [09:23:52] Barry Leiba, Right. [09:24:11] Barry Leiba, Actually, keep on saying more intelligent things, and I'll claim that was what I meant. [09:24:19] john klensin: many mixed feelings. This can turn into bureaucracy. [09:24:34] dc: :-) [09:24:40] g.e.montenegro leaves the room [09:25:30] john klensin: the app area meeting used to be to get through documents. now it's just 2 1/2 hours of discussion. [09:25:58] john klensin: if we've got a process problem, let's fix the process problem. [09:26:18] (denying what Pete and not Don Quixote said) [09:26:19] EKR leaves the room [09:26:29] alexey: Not clear where we've landed. [09:27:08] alexey: would like to hear on apps-discuss how else we can deal with the individual drafts. [09:27:56] Randall Gellens leaves the room [09:27:59] dave crocker: understand that the current model is too hard on the ADs. maybe we should flesh that out and the community can work on a solution. [09:28:14] alexey: maybe we can bring more detail to the list. [09:28:40] Julian leaves the room: Computer went to sleep [09:28:43] Yves Lafon leaves the room [09:28:48] cyrus leaves the room [09:28:54] sm leaves the room [09:28:57] tlyu leaves the room [09:28:59] end of session [09:29:00] Atarashi Yoshifumi leaves the room [09:29:02] mtcarrasco leaves the room [09:29:04] sgerdes leaves the room [09:29:08] Haibin leaves the room: Computer went to sleep [09:29:18] yone leaves the room [09:29:28] John C Klensin leaves the room [09:29:35] Cary leaves the room [09:29:49] fujiwara leaves the room [09:30:25] ANd now the audio becomes interesting. :-) [09:30:40] hehe [09:30:40] marc.blanchet.qc leaves the room [09:31:00] healthyao2000 leaves the room [09:31:11] stpeter leaves the room: Logged out [09:31:38] Joe Hildebrand leaves the room: Disconnected. [09:32:13] Joe Hildebrand joins the room [09:32:24] resnick leaves the room [09:32:43] paf leaves the room [09:33:18] Barry Leiba leaves the room [09:36:53] Joe Hildebrand leaves the room: Disconnected. [09:37:15] chryss@jabber.org/Meebo joins the room [09:37:41] chryss@jabber.org/Meebo leaves the room [09:38:56] ray_atarashi leaves the room [09:39:11] chris waigl joins the room [09:44:01] tonyhansen leaves the room [09:44:17] chris waigl leaves the room [09:45:05] tonyhansen joins the room [09:45:07] Randall Gellens joins the room [09:54:40] Mark Nottingham joins the room [09:54:57] Mark Nottingham leaves the room [10:05:09] Mark Nottingham joins the room [10:21:16] Yves Lafon joins the room [10:27:27] Randall Gellens leaves the room [10:38:09] Yves Lafon leaves the room [10:42:22] Mark Nottingham leaves the room [10:44:22] martin.thomson joins the room [10:53:28] martin.thomson leaves the room [10:58:56] martin.thomson joins the room [11:04:10] giomarti@cisco.com joins the room [11:05:22] giomarti@cisco.com leaves the room [11:06:39] ray_atarashi joins the room [11:06:55] ray_atarashi leaves the room [11:07:32] paf joins the room [11:07:50] tonyhansen leaves the room [11:07:54] tonyhansen joins the room [11:09:42] stpeter joins the room [11:10:33] paf leaves the room [11:13:02] stpeter leaves the room: Logged out [11:13:45] stpeter joins the room [11:17:43] tonyhansen leaves the room [11:21:23] martin.thomson leaves the room [11:30:02] Joe Hildebrand joins the room [11:33:14] Joe Hildebrand leaves the room [12:14:53] stpeter leaves the room: Disconnected: connection closed [12:32:48] stpeter joins the room [13:13:18] JLCjohn leaves the room [13:43:11] stpeter leaves the room: Disconnected: connection closed [13:51:05] Dave Cridland leaves the room: offline [14:09:38] stpeter joins the room [14:36:01] stpeter leaves the room [18:20:35] Kev leaves the room