IETF
homenet@jabber.ietf.org
Tuesday, 31 July 2012< ^ >
Room Configuration

GMT+0
[14:44:04] Christoffer Holmstedt joins the room
[14:50:56] Michael Thomas joins the room
[15:31:54] Michael Thomas leaves the room
[15:39:25] Michael Thomas joins the room
[15:46:15] Michael Thomas leaves the room
[15:46:16] Michael Thomas joins the room
[16:36:14] Michael Thomas leaves the room
[16:55:57] Michael Thomas joins the room
[17:13:58] Christoffer Holmstedt leaves the room
[17:45:24] Michael Thomas leaves the room
[18:32:01] Michael Thomas joins the room
[20:29:32] Michael Thomas leaves the room
[20:29:39] Michael Thomas joins the room
[20:37:19] Michael Thomas leaves the room
[20:37:23] Michael Thomas joins the room
[21:12:35] Michael Thomas leaves the room
[21:12:35] Michael Thomas joins the room
[21:13:23] yi.jiazi@gmail.com joins the room
[21:13:46] yi.jiazi@gmail.com leaves the room
[21:15:50] Michael Thomas leaves the room
[21:15:50] Michael Thomas joins the room
[21:22:28] Michael Thomas leaves the room
[21:22:28] Michael Thomas joins the room
[21:31:07] wmtownsley joins the room
[21:46:33] Markus Stenberg joins the room
[22:09:17] yi.jiazi@gmail.com joins the room
[22:11:14] cabo joins the room
[22:12:20] Michael Thomas leaves the room
[22:14:48] yi.jiazi@gmail.com leaves the room
[22:15:58] dyork joins the room
[22:16:16] yi.jiazi@gmail.com joins the room
[22:17:20] <dyork> Homenet meeting at IETF 84 - Vancouver
[22:17:25] <dyork> Agenda: http://tools.ietf.org/wg/homenet/agenda?item=agenda-84-homenet.html
[22:19:26] sandoche joins the room
[22:21:07] Michael Thomas joins the room
[22:22:03] Satoru Kanno joins the room
[22:23:48] <dyork> Hi. I'll be jabber-scribing and relaying
[22:23:53] <dyork> Do we have remote attendees?
[22:24:04] <Markus Stenberg> yes
[22:24:13] Christoffer Holmstedt joins the room
[22:24:16] Atarashi Yoshifumi joins the room
[22:24:31] <dyork> Great. Please just put "mic:" at the beginning of anything you want to relay and I'll be glad to do so. I'm near a mic
[22:24:49] <wmtownsley> Dan, thank you for being our scribe.
[22:24:59] <wmtownsley> Thomas Clausen is our note taker.
[22:25:02] jerrycwt joins the room
[22:25:13] cgriffiths joins the room
[22:25:24] <dyork> In the Intro slides: http://www.ietf.org/proceedings/84/slides/slides-84-homenet-0.pdf
[22:25:40] <dyork> Slide 3: Administrivia
[22:25:40] gigix73 joins the room
[22:25:59] <dyork> Slide 4: Agenda Bash
[22:26:28] geordie bilkey joins the room
[22:26:34] geordie bilkey leaves the room
[22:26:36] Wassim M. Haddad joins the room
[22:26:51] Geordie Bilkey joins the room
[22:26:58] <dyork> mark - you're welcome
[22:27:07] Christoffer Holmstedt leaves the room
[22:28:35] <dyork> Next up - Tim Chown on Homenet architecture. Slides: http://www.ietf.org/proceedings/84/slides/slides-84-homenet-1.pdf
[22:28:41] <Michael Thomas> is the audio streaming working?
[22:28:52] Suzanne joins the room
[22:28:55] Andrew Sullivan joins the room
[22:28:58] <dyork> Don't know. :-(
[22:29:10] christoffer.holmstedt joins the room
[22:29:13] <Markus Stenberg> yep it is
[22:29:22] <dyork> Audio should be at: http://ietf84streaming.dnsalias.net/ietf/ietf846.m3u
[22:29:33] <Suzanne> And we're off....arch doc updates in brief, more detailed discussion of some changes comes later
[22:29:49] <dyork> Slide 2 - Updates from -02
[22:30:20] <dyork> Slide 3 - Next steps?
[22:31:12] Tina TSOU joins the room
[22:32:07] <dyork> Tim Chown asking if current structure and content captures the issues and list discussions sufficiently well
[22:32:14] <dyork> Michael Richardson at mic
[22:34:22] <dyork> Kerry Lynn at mic
[22:34:51] <dyork> Kerry believes document should be more requirements-based
[22:35:33] dmitry.anipko joins the room
[22:36:15] <Suzanne> Mark: hearing it's not fully baked, needs another rev to tighten the prose
[22:37:14] <dyork> Hmmm... slides don't seem to be available
[22:37:47] <Suzanne> Naming and Service Discovery, per Stuart Cheshire-- not from the list, slides apparently just uploaded
[22:38:09] Michael Richardson joins the room
[22:38:15] timc joins the room
[22:38:23] <dyork> Here they are: http://www.ietf.org/proceedings/84/slides/slides-84-homenet-2.pdf
[22:38:33] <dyork> Slide 3
[22:38:41] <dyork> Single Answer vs Multiple Answers
[22:38:44] <Michael Richardson> if we dn't have a jabber->mic, person, I can volunteer. is anyone remote?
[22:38:54] <dyork> I'm covering it
[22:38:58] <Michael Richardson> good.
[22:38:59] <dyork> but glad to have a backup
[22:38:59] <Suzanne> with backup from me :)
[22:39:08] <dyork> cool! Multiple backups
[22:39:15] <dyork> Doesn't often happen! :-)
[22:39:25] jinmei joins the room
[22:39:35] <dyork> Slide 4: Single Answer vs Multiple Answers
[22:39:50] <Michael Thomas> anybody have the url of the slides?
[22:39:54] <dyork> Slide 5: Naming Comes in Many Guises
[22:39:55] <Suzanne> host vs. service....distinction is blurry, not obvious
[22:39:57] <dyork> http://www.ietf.org/proceedings/84/slides/slides-84-homenet-2.pdf
[22:40:05] <Michael Thomas> thx
[22:40:30] <dyork> Slide 6
[22:40:53] <dyork> Slide 6: Naming Comes in Many Guises (2nd slide with that title)
[22:41:11] <dyork> Slide 7: multicast vs repository
[22:42:13] <dyork> If you're remote, please put "mic:" at the beginning of anything you'd like us to relay. We've got several of us who will do so.
[22:43:13] <dyork> Slide 8: Local vs. Global
[22:44:12] tsuichi joins the room
[22:45:07] <dyork> Last slide: Naming & Service Discovery
[22:45:43] <dyork> Lee Howard at mic
[22:45:48] <Suzanne> "More similar than different"
[22:46:12] ogud joins the room
[22:48:13] <dyork> Ralph Droms at mic
[22:50:12] <dyork> Anders Brandt at mic
[22:50:51] <dyork> David Oran at mic
[22:52:22] <Suzanne> Ray Bellis speaking "no hat"
[22:52:48] <dyork> Lee Howard back at mic
[22:53:18] <dyork> Kerry Lynn at mic
[22:54:10] <dyork> Anders Brandt at mic
[22:54:37] <dyork> (he's next - Stuart replying now to Kerry)
[22:55:26] <dyork> Ralph Droms at mic
[22:56:40] <dyork> Andrew Mcgregor at mic
[22:56:48] yi.jiazi@gmail.com leaves the room
[22:57:23] yi.jiazi@gmail.com joins the room
[22:57:58] <dyork> Andrews Sullivan at mic
[22:58:15] <Suzanne> Stuart: example of naming complexity that's not user-friendly, two names for same printer from local and remote access
[22:58:20] gigix73 leaves the room
[22:58:58] <dyork> Toerless Eckert at mic
[23:00:37] <Suzanne> Stuart: further discussion, examples of names where "aliases" or identifiable multiple names for one resource for different scopes would be useful
[23:01:44] <dyork> Ralph Droms at mic
[23:02:26] <Michael Thomas> mic: if you posit a global name as a cname for a .local name, why use the .local at all?
[23:03:17] <dyork> k
[23:03:31] <dyork> thomasm - what is your full name?
[23:03:32] gigix73 joins the room
[23:03:36] <Michael Thomas> mike thomas
[23:03:42] <dyork> Jonne Sonnen at mic
[23:03:44] <dyork> Thanks
[23:05:25] <dyork> anders brandt at mic
[23:06:17] <dyork> jonne back
[23:06:30] <dyork> anders
[23:06:30] sujing joins the room
[23:07:17] herve.prigent joins the room
[23:08:12] <Suzanne> Dave Thaler
[23:08:17] <dyork> Dave Thaler
[23:08:31] <dyork> Thx, suzworldwide
[23:09:47] Michael Thomas leaves the room
[23:09:47] Michael Thomas joins the room
[23:10:55] Suzanne leaves the room
[23:10:55] <dyork> Kerry Lynn at mic
[23:10:55] dyork leaves the room
[23:10:56] herve.prigent leaves the room
[23:11:04] dyork joins the room
[23:11:21] <Michael Thomas> who is speaking?
[23:11:24] suzworldwide joins the room
[23:11:37] <dyork> http://www.ietf.org/proceedings/84/slides/slides-84-homenet-3.pdf
[23:11:51] <dyork> Daniel Migault
[23:11:52] <suzworldwide> II Naming Architecture Requirements
[23:12:02] <dyork> Anyone else having jabber issues?
[23:12:08] <suzworldwide> III Naming Delegaton Architecture
[23:12:13] <Michael Thomas> i'm ok here
[23:12:49] <suzworldwide> IV Setting Naming Delegation
[23:12:58] <dyork> We're jumping through the slides quickly
[23:13:04] <Michael Thomas> yeah, i gues
[23:13:06] <dyork> http://www.ietf.org/proceedings/84/slides/slides-84-homenet-3.pdf
[23:13:15] <dyork> We're up to slide 9
[23:13:24] <dyork> (these slides are at least numbered)
[23:13:48] <dyork> slide 10
[23:14:26] <suzworldwide> V DHCP Exchanges
[23:14:34] <dyork> slide 11
[23:14:39] <dyork> What to do next
[23:14:52] <dyork> Now slides: http://www.ietf.org/proceedings/84/slides/slides-84-homenet-4.pdf
[23:15:08] <dyork> slide 1 Table of Contents
[23:15:20] <dyork> Front End Naming Architecture
[23:15:31] <dyork> Slide 2 - Introduction
[23:15:58] <dyork> Draft being discussed is http://tools.ietf.org/html/draft-mglt-front-end-naming-delegation-00
[23:16:20] <dyork> Slide 4
[23:17:00] <dyork> Slide 5
[23:17:02] <dyork> Slide 6
[23:17:16] <dyork> III. Setting Front End Naming Delegation
[23:18:06] <Michael Thomas> when are they going to take questions?
[23:18:17] <dyork> after these two presos
[23:18:21] <Michael Thomas> k thx
[23:18:22] <dyork> Next slide
[23:18:28] <dyork> IV DHCP Exchange
[23:19:06] <dyork> Last slide - what to do next
[23:19:12] <dyork> Big queue forming
[23:19:20] <suzworldwide> Lee Howard first
[23:19:27] <dyork> Lee Howard at mic
[23:19:33] <dyork> long queue forming
[23:19:42] <dyork> Mike - you want to comment? (i.e. should I get in line)
[23:19:56] <Michael Thomas> let me hear the q's first...
[23:20:17] <Markus Stenberg> .. I don't think encoding of options is relevant, but blocking yet another port to CPE from wide world leads to worse customer experience
[23:20:52] <dyork> Mark T. pointing out that the preso is for a managed CPE
[23:20:53] <dyork> Stuart Cheshire at mic
[23:22:11] <dyork> Stuart sees value in auto-configuration of device
[23:23:04] <dyork> Erik Kline at mic
[23:23:40] <suzworldwide> Erik asks how updating works for this architecture
[23:24:03] <Michael Thomas> mic: the big problem here is the implicit tight coupling between the ISP and the CPE by using DHCP. What if I want to want to use another service where DHCP isn't an option, like say google. If this is decoupled from DHCP, it would be a lot more reasonable
[23:24:19] <dyork> Mike - the chairs have unfortunately cut the line
[23:24:40] <dyork> Don't know if they'll let me squeeze in a remote question
[23:24:45] <suzworldwide> Chris Griffiths
[23:24:51] <dyork> Chris Griffiths at mic
[23:25:07] <Michael Thomas> i would appreciate it if it doesn't get brought up
[23:25:18] <Markus Stenberg> .. and multiple ISPs, ISP changes..
[23:25:19] <dyork> Mark Andrews at mic
[23:25:23] <Michael Thomas> yes
[23:25:25] <Markus Stenberg> it sounds like draft written by ISP people
[23:25:31] <Michael Thomas> it was
[23:25:44] <dyork> yes... ISP with managed CPE
[23:28:10] <dyork> Toerless Eckert at mic
[23:28:50] <dyork> Peter Koch at mic
[23:29:35] <suzworldwide> Peter notes layer 9 implications to naming scopes
[23:29:38] <dyork> Andrew Sullivan at mic
[23:31:36] paul.beaumont joins the room
[23:31:45] <Michael Richardson> the marketing/sales department of most ISPs is desperately looking for ways of locking customers in so that they can reduce churn. Maybe the DNS guys at the ISP don't like it... but..
[23:32:06] <dyork> Next preso - Tim Chown back up
[23:32:13] <Michael Richardson> and if you want to use google, then buy a CPE that supports google. We use DHCP because we want it automated.
[23:32:15] <dyork> http://www.ietf.org/proceedings/84/slides/slides-84-homenet-5.pdf
[23:32:39] <suzworldwide> name service design principles
[23:32:48] <dyork> Slide 2 - Name service design
[23:33:02] <dyork> Slide 3- Fundamentals
[23:33:53] <dyork> Slide 4 - Namespace
[23:35:48] <dyork> Dave Thaler at mic
[23:36:02] <dyork> Okay... big line forming
[23:38:50] <dyork> Robby Simpson at mic
[23:39:11] arifumi joins the room
[23:39:15] <Andrew Sullivan> People keep saying "CNAME" here when I think what they mean is "SRV".
[23:39:32] <dyork> Kerry Lynn at mic
[23:39:41] jweil joins the room
[23:39:53] <Michael Thomas> i think that cname was brought into the discussion from stewart
[23:40:49] <dyork> Anders Brandt at mic
[23:41:03] Wassim M. Haddad leaves the room
[23:41:29] <dyork> Stuart Cheshire at mic
[23:42:16] <dyork> "we need a way to reach local things when the global connection is down"
[23:42:51] <Andrew Sullivan> This is what split horizon DNS is for. It's awful, but it already does this.
[23:43:16] <Michael Thomas> what is .local if not split horizon?
[23:43:17] <Michael Richardson> split-horizon DNS already failed for every mobile node.
[23:43:33] <dyork> Mark Andrews at mic
[23:43:37] <Andrew Sullivan> @Michael Richardson: yes.
[23:43:43] <Michael Richardson> .local does not reuse example.com internally and externally.
[23:44:36] <Michael Thomas> yes, but it causes different answers depending on scope. that's the salient thing about split horizon imo
[23:44:42] <dyork> Mark Andrews making point that a name server can be inside the household... and do DNSSEC signing of .local zone
[23:44:46] <Michael Richardson> Andrew Sullivan can you post to the list an example of what you think mDNS would say, exactly, that would involve SRV? I think I agree, but I need an example.
[23:45:19] <dyork> (?) at mic
[23:45:25] <Michael Thomas> no
[23:45:36] <Michael Thomas> oh, nevermind :)
[23:45:56] <dyork> No nametag on speaker at mic and couldn't get his name.
[23:46:22] <Andrew Sullivan> I'll try to work up a convincing example. My point was just that the idea here seems to be that we need to have some way of saying "this service you know by $name1 is actually globally known as $name2", and that's something SRV records already can do.
[23:46:36] <dyork> Next - Independent operation
[23:47:20] <dyork> Next - Some security-related issues
[23:47:47] <cgriffiths> SRV seems like a good example to handle not only service and port information, but also priority and weight for naming
[23:48:23] <cgriffiths> in order to deal with the internal vs global naming
[23:48:33] <dyork> Lee Howard at mic
[23:48:56] <dyork> Anders Brandt at mic
[23:49:35] <dyork> Kerry Lynn at mic
[23:50:20] <suzworldwide> Olafur Gudnundsson
[23:50:26] <dyork> thx
[23:50:53] <dyork> Peter Koch at mic
[23:50:57] <suzworldwide> Gudmundsson rather, sorry Olafur
[23:51:10] ogud leaves the room
[23:51:32] <dyork> Dave Thaler at mic
[23:52:39] tsuichi leaves the room
[23:53:04] <dyork> Andrew Sullivan at mic
[23:53:35] ogud joins the room
[23:54:02] jerrycwt leaves the room
[23:54:19] <dmitry.anipko> comment to the mic queue: one way to solve ambiguity of local names is to have a strong ID of home network (example - self signed cert) to which a host can reliably remember and let user effectively choose the context of the name (selection of the context may affect what actual name is resolved on the wire).
[23:54:25] <dyork> k
[23:54:53] <dyork> stuart
[23:56:08] <wmtownsley> "Hexadecimal vomit is not a good user experience"
[23:56:17] <Tina TSOU> no sound?
[23:56:20] <dmitry.anipko> no sound
[23:56:27] <Markus Stenberg> sound went poof 2 min ago
[23:56:30] <Michael Thomas> ok, it's not just me
[23:56:33] <dyork> Evan Hunt at mic
[23:56:35] gigix73 leaves the room
[23:56:47] Satoru Kanno leaves the room
[23:58:30] <dyork> Bob Hinden at mic
[23:58:46] <dmitry.anipko> we can't hear the mic
[23:59:09] <dyork> Can you hear the chairs or Tim?
[23:59:13] Michael Thomas leaves the room
[23:59:32] <Markus Stenberg> nah, whole stream's dead
[23:59:35] <Markus Stenberg> can't even connect to it
[23:59:37] <dmitry.anipko> same here
[23:59:40] <dyork> Hmmm
[23:59:41] Michael Thomas joins the room
[23:59:51] <dyork> Dang... I relayed your question, dmitry :-)
Powered by ejabberd Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!