IETF
homenet@jabber.ietf.org
Tuesday, 15 November 2011< ^ >
Room Configuration

GMT+0
[00:28:58] mcr joins the room
[00:30:39] <mcr> my thanks to those who provided their slides in PDF.
[00:36:18] <mcr> http://junk.sandelman.ca/junk/ietf82/
[00:36:22] <mcr> contains everything in PDF.
[00:56:08] jlcJohn joins the room
[00:58:07] yaovct joins the room
[00:59:19] <mcr> it's very quiet?
[01:00:06] <mcr> It was noisier before.
[01:00:57] <mcr> yes, I frobbed knobs.
[01:01:10] <mcr> who is local? who is remote?
[01:01:13] <mcr> (I am remote)
[01:01:39] <babongo> (remote)
[01:01:52] Ole Troan joins the room
[01:01:58] ray_atarashi joins the room
[01:02:01] arifumi joins the room
[01:02:14] <Ole Troan> I'm now the jabber relay, volunteer.
[01:02:19] <Ole Troan> anyone remote?
[01:02:26] <Ole Troan> agenda bash
[01:02:49] monden joins the room
[01:03:12] Colman Ho joins the room
[01:03:19] dmitry.anipko joins the room
[01:03:20] monden leaves the room
[01:04:01] dudisaki joins the room
[01:04:53] <Ole Troan> update from the interim
[01:05:00] Atarashi Yoshifumi joins the room
[01:05:04] becarpenter joins the room
[01:05:54] tatsuji joins the room
[01:06:00] naptee joins the room
[01:06:14] <mcr> Ole, babongo, and I are remote.
[01:06:28] <mcr> maybe others.
[01:06:32] <Ole Troan> OK, let me give some more detail then
[01:06:36] monden joins the room
[01:06:40] <Ole Troan> Tim on the homenet architecture draft. slide 1
[01:06:55] dmitry.anipko leaves the room
[01:07:02] <Ole Troan> slide 2. where we're at
[01:07:06] Lee Howard joins the room
[01:07:07] <mcr> ietf82/homenet-7.pdf <file:///home/mcr/Desktop/Downloads/ietf82/homenet-7.pdf>
[01:07:14] resnick joins the room
[01:07:25] <mcr> http://junk.sandelman.ca/junk/ietf82/homenet-7.pdf <file:///home/mcr/Desktop/Downloads/ietf82/homenet-7.pdf>
[01:07:44] Chris Griffiths joins the room
[01:08:00] <Ole Troan> slide 3
[01:08:31] <Ole Troan> all presentations here: https://datatracker.ietf.org/meeting/82/materials.html#wg-homenet
[01:08:37] <Ole Troan> slide 4
[01:09:20] TACHIBANA toshio joins the room
[01:09:22] <Ole Troan> *5
[01:09:41] tsavo_work@jabber.org/Meebo joins the room
[01:10:11] h-suzuki joins the room
[01:10:20] naito.kengo@jabber.org joins the room
[01:10:39] <mcr> Ole: but the important thing is that it's homenet-7.
[01:10:41] <Ole Troan> *6
[01:10:50] <Ole Troan> mcr: ack
[01:11:13] ray joins the room
[01:12:12] behcet.sarikaya joins the room
[01:12:42] <Ole Troan> *7
[01:12:46] <resnick> Can we speed this up? There is nothing new here.
[01:13:03] <resnick> Everyone who is paying attention knows this already.
[01:13:13] <mcr> +1
[01:13:48] <Ole Troan> *8
[01:14:18] <Ole Troan> *9
[01:14:29] sftcd joins the room
[01:14:48] <Ole Troan> *10
[01:15:11] <Ole Troan> *11
[01:15:29] <Ole Troan> Switching speaker to Jari
[01:15:55] Andrew Sullivan joins the room
[01:15:57] <Ole Troan> *12
[01:17:02] angga joins the room
[01:17:08] kawashimam joins the room
[01:17:42] <Ole Troan> Q from Erik Nordmark
[01:17:57] vikas.sarawat joins the room
[01:18:32] Suz joins the room
[01:19:34] <Ole Troan> Q: Lorenzo
[01:20:03] Suz leaves the room
[01:20:06] <mcr> mic: while few people want to break their IPv4, there are lots of networks that *do not yet exist* (except at HouseArko), which might not need IPv4 at all.
[01:20:47] <resnick> Ralph is very irresponsible. :-D
[01:20:50] Suz joins the room
[01:20:58] <Ole Troan> mcr: do you want that stated on the mike?
[01:21:12] Bing Liu joins the room
[01:21:24] <mcr> mic: "homenet admin"??? you mean the 12 year old!!!!
[01:21:27] <mcr> yes...
[01:21:55] <Ole Troan> too long a queue… ;-)
[01:21:55] <ray> mcharlesr - 4 in the line for the mic already....
[01:22:00] <mcr> okay.
[01:23:38] <mcr> http://www.ietf.org/proceedings/82/slides/6man-2.pdf slide 5, shows a situation *TODAY* where customers plug things in randomly.... if modem was router...
[01:25:12] lllmartini joins the room
[01:25:45] pselkirk joins the room
[01:25:58] antoin joins the room
[01:26:09] <Ole Troan> Ralph interjects comment. after Mark talking
[01:26:23] <Ole Troan> Peter Lothberg:
[01:27:26] <mcr> Peter really really really really says it well!!!!!
[01:27:32] <antoin> louder please, the mic in the room is hardly understandable
[01:27:46] <ray> (noted)
[01:29:11] <Ole Troan> Alain Durand at the mike
[01:29:33] Dave Thaler joins the room
[01:29:59] <mcr> nothing speaks IPv4. It speaks HTTP.
[01:30:05] <Dave Thaler> how long is line cut for? this topic? rest of slide? rest of preso? I have a comment later on this slide.
[01:30:15] <ray> just for this slide, Dave
[01:30:22] naptee leaves the room
[01:30:28] <Dave Thaler> so I can't raise another issue on the part of the slide Jari hasn't covered yet?
[01:30:36] <ray> sure, that'll be OK
[01:31:01] naptee joins the room
[01:31:03] <Dave Thaler> (ok more comment/suggestion than issue)
[01:31:16] Olafur joins the room
[01:31:41] naptee leaves the room
[01:32:34] <Ole Troan> Lee Howard at the mike
[01:32:38] naptee joins the room
[01:33:00] <Ole Troan> Peter Lothberg
[01:33:19] <Ole Troan> * still discussion on slide 12 dual stack homenets 8
[01:34:28] <Ole Troan> Dave Thaler:
[01:35:31] <Ole Troan> Erik Nordmark:
[01:36:14] <Ole Troan> Dave: "make IPv4 and IPv6 topologies congruent. treat them as ships in the night, but make them follow the same shipping lanes".
[01:36:52] evyncke joins the room
[01:37:25] <Ole Troan> Peter
[01:37:43] <Ole Troan> slide 13
[01:42:07] <Ole Troan> Anders ?:
[01:42:35] monden leaves the room
[01:42:37] <Ole Troan> Tom Herbst
[01:44:13] satoru.matsushima joins the room
[01:44:56] <Ole Troan> Brian Carpenter
[01:45:42] monden joins the room
[01:46:05] monden leaves the room
[01:46:36] <Ole Troan> slide 14
[01:47:47] Warren Harrop joins the room
[01:49:23] <Ole Troan> Lee at the mike
[01:50:43] <Ole Troan> Marc Blanchet
[01:50:52] <mcr> we need to keep the addresses somewhat stable (including the ULA ones!!!) so that we can make the security stuff work.
[01:51:28] <Ole Troan> mcr: that's clearly the goal
[01:51:33] <mcr> the ISP can renumber the /56 above, but the 56-64 could still remain constant.
[01:51:57] <mcr> naming and service location can't inform the security...
[01:51:58] <Lee Howard> that's another good example. the requirement is that security policies not break due to renumbering. One way is not to renumber. Another is to make the implementation of security policies be resilient to changes.
[01:52:01] <Ole Troan> slide 15
[01:52:32] <Lee Howard> s/resilient/responsive
[01:52:34] <mcr> Andrew's hosts have stable link-local addresses to find the printer, and his printer has a stable address for him to determine if he is allowed to print.
[01:52:43] <Ole Troan> right, so "subnet-id" stability is important and security policies should be applied to that as opposed to varible ISP prefix
[01:52:54] <Lee Howard> that would work, too
[01:53:55] <Ole Troan> Peter L
[01:54:05] <Lee Howard> or, prefix delegation/assignment could be algorithmic, so that any given interface would get the same relative prefix for a given place in the topology
[01:54:38] tomek joins the room
[01:54:42] <mcr> Lee, I agree. the security could adapt to the renumbering. Years ago we tried to wave DNSSEC at the security-by-numbering problem. I think we can continue to do this, but there is a bootstrap problem here.
[01:54:47] jweil joins the room
[01:55:02] <Ole Troan> Lee: that's an interesting idea
[01:55:34] <Ole Troan> Lorenzo @mike
[01:55:37] <Lee Howard> To be absolutely clear (and I realize the room has moved on): I'm not even necessarily opposed to the solution; I am opposed to listing the solution as the requirement. Make clearer requirements.
[01:56:07] resnick leaves the room
[01:56:12] <Suz> I like it, *if* it turns out to be possible to make reasonable assumptions about how stable "a given place in the topology" actually is. Could be a cool heuristic to have in some situations, but enough to be worth the work?
[01:56:25] <mcr> Ole, Lee, the Arch(1) mechanism burns prefixes into flash on each device. The Arch(2) burns them into flash on "master" PD DHCP device. So, picking consistent addresses is not hard.
[01:57:23] <Ole Troan> yes, you can use persistent storage, but stability cannot be expected when the topology changes
[01:57:53] <mcr> we have guest + n*private, (HouseBaker), but I agree that this isn't Grandma.
[01:58:47] <Ole Troan> Erik N:
[01:59:36] <Lee Howard> slide 16
[02:00:20] <Ole Troan> 17
[02:00:24] <Ole Troan> 18
[02:00:51] <mcr> I think we should repeat the "All other aspects of multihoming are out of scope" at the beginning of each meeting.
[02:02:32] <Lee Howard> slide 19
[02:02:48] <Ole Troan> 20
[02:02:53] <Ole Troan> ups, 19
[02:02:59] <Lee Howard> Dave Thaler at mike
[02:03:10] <Ole Troan> 20
[02:04:11] <mcr> www.we-love-ula.com <http://www.we-love-ula.com>
[02:04:44] <Ole Troan> Brian Carpenter
[02:05:24] <mcr> what about a network where parents are prevents from spying on their children?...
[02:05:33] <mcr> s/prevents/prevented/
[02:05:54] <Ole Troan> crypto? why is that a network layer problem?
[02:06:02] <ray> mine aren't spied on, but they can't get out without going through my squid proxy with a site whitelist....
[02:06:08] <Lee Howard> who said it was network layer?
[02:06:10] <Ole Troan> next presentation http://www.ietf.org/proceedings/82/slides/homenet-9.pptx
[02:06:26] <mcr> Ole: multicast DNS can tell the entire network what sites are being looked up....
[02:06:30] <Ole Troan> Lee: isn't homenet doing stuff at L3?
[02:06:30] <Lee Howard> Cary Lynn?
[02:06:51] <Lee Howard> I don't think service discovery, naming, security policy are necessarily L3.
[02:07:19] <babongo> Kerry Lynn
[02:07:22] Lee Howard leaves the room
[02:07:41] <mcr> becarpenter: is there an issue with privacy (extensions) in the subnet-id part of the address? Are we leaking info there of significance?
[02:08:18] <ray> Erik's presentation (after Ole) is at http://www.ietf.org/proceedings/82/slides/homenet-0.pptx
[02:09:04] <becarpenter> If the IID is fixed, traffic can be tied back to an individual host (so you can answer questions like - who has been using Facebook in the last 5 minutes?)
[02:09:35] <becarpenter> Privacy address were designed to avoid this, but that means that the IID keeps changng
[02:10:15] <becarpenter> And if you have a separate subnet for the kids' bedroom...
[02:10:24] <mcr> yes, but is it important to know if facebook was accessed from a particular subnet.
[02:10:54] <mcr> ah, yes... so it does matter. Right now, with bridged wifi, you can't tell if facebook was accessed via wired or wireless.
[02:10:58] <becarpenter> 'Honey, one of the kids is using Facebook again'
[02:12:09] <mcr> how is it different from, "Honey, someone just used facebook. It wasn't me. If it wasn't you...?"
[02:12:26] <evyncke> Or even worse: honey, why were you accessing our NAS from a different city than Taipei? Are you not at te IETF? would you be at your mistress' home? :)
[02:12:46] <mcr> evyncke, that's why IETF needs a NEMO server.
[02:12:55] <ray> ted lemon at the mic
[02:14:04] <becarpenter> @mcr: true, it refines traffic analysis but you're correct, a savvy parent or spouse can do quite a bit with bridged traffic
[02:14:52] <mcr> evyncke, becarpenter, more to the point, Taipei vs HouseMistress, or child bedroom vs dining room, is encoded in the prefix, not in the lower-64, anyway.
[02:15:16] <evyncke> if you use privacy extension you are usually fine
[02:15:31] Lee Howard joins the room
[02:15:38] <ray> fred baker speaking
[02:15:48] <evyncke> Of course if you also use bonjour to adv your name & service even privacy extension will not help :-)
[02:15:51] <becarpenter> Correct, as soon as you have subnets the prefix is quite informative
[02:16:13] <mcr> "under what conditions is the invalidation valid?"
[02:17:03] <antoin> louder please, I could not hear Ted
[02:17:16] <mcr> +1
[02:17:25] monden joins the room
[02:17:59] <Ole Troan> http://www.ietf.org/proceedings/82/slides/homenet-0.pptx
[02:18:02] <Ole Troan> slide 2
[02:18:26] <ray> numbered #3
[02:18:52] <ray> 4
[02:18:53] <Ole Troan> 4
[02:20:19] <mcr> guest networks come in both flavours: ACLs (often layer-2) and seperate ESSID. Sometimes both.
[02:20:28] <Ole Troan> 5
[02:22:04] <Lee Howard> 6
[02:24:24] <Lee Howard> 7
[02:24:57] <ray> Fred's slides are at http://www.ietf.org/proceedings/82/slides/homenet-8.pptx
[02:26:15] <becarpenter> And where is the non-PPTX version?
[02:27:27] <mcr> becarpenter: http://junk.sandelman.ca/junk/ietf82/ <http://junk.sandelman.ca/junk/ietf82/>
[02:27:54] <becarpenter> Cool!!!
[02:28:16] <mcr> it's what I did while waiting for the session to start.
[02:28:27] <Lee Howard> NAMES please!
[02:28:36] <mcr> it's Andrew McGregor.
[02:28:49] <antoin> louder please
[02:28:53] evyncke leaves the room
[02:29:44] <Lee Howard> "louder" relayed
[02:30:32] <ray> anyone else having trouble hearing the rear mic, or just Antoin?
[02:30:45] <antoin> thanx, it helped a bit, but I notice that the mikes in the room are a lot less loud than the ones on the stage
[02:31:35] <antoin> so I have to turn up the volume when someone asks a question from the room
[02:31:47] <Suz> all of the mics sound a little faint, in the room is worse
[02:32:15] evyncke joins the room
[02:32:44] <Ole Troan> Fred Baker presentation
[02:32:53] <Ole Troan> http://www.ietf.org/proceedings/82/slides/homenet-8.pptx
[02:33:01] <mcr> I don't have problems with any of the speakers. Yes, the main speaker is a bit high, but I'm not wearing headphones.
[02:35:35] <ray> "tree network" slide
[02:36:08] <mcr> mic: fred's basic cases do not include Erik's basic cases, specifically, the case where it does a much better job, which is the chained case of VoIP/Backup/ISP-router case.
[02:36:10] <ray> "multihoming"
[02:36:25] <ray> Ole - please relay Michael's comment
[02:37:22] <mcr> (er, Fred's protocol does a better job than Erik's for the case Erik says is typical)
[02:37:48] <evyncke> agree
[02:38:27] <mcr> (well, I apologize for personalizing the protocols.)
[02:39:50] <mcr> ray: it would be nice if the slides that go into the archive could include the name of Fred's draft on the first slide.
[02:39:58] <Lee Howard> Lothberg at the mike
[02:40:09] <mcr> draft-baker-homenet-prefix-assignment-00 is it from the agenda.
[02:40:46] <Dave Thaler> since I'm not allowed to get up... there seems to be two separate classes of scenario: topologies that work with IPv4 home NATs today and topologies that don't (Erik's point in last preso).
[02:40:48] <Lee Howard> Jari Arkko at the mike
[02:40:55] Keiichi SHIMA joins the room
[02:41:00] <Dave Thaler> for the ones that IPv4 works, this seems the same as Erik's.
[02:41:11] Keiichi SHIMA leaves the room
[02:41:14] <Ole Troan> Dave: it isn't that they don't work. but there are proposals that are _restricted_ to only DAG topologies
[02:41:35] <Dave Thaler> yeah, it would help to frame the discussion that way
[02:41:51] <Dave Thaler> the "simple" scenarios and the "advanced" scenarios (or some other terms)
[02:42:32] <Ole Troan> agree, and that's back to Lorenzo's comment. unless we can find a way to evolve from simple to advanced, we might as well just go with advanced directly...
[02:42:45] <Dave Thaler> the "advanced" (or whatever) scenarios aren't about adding IPv6 to an IPv4 home net, they're about rearchitecting it (or managing it, if there's an admin of some type) for both IPv4 and IPv6
[02:42:45] <Lee Howard> Pascale Thubert, I think
[02:43:07] <evyncke> indeed this is Pascal (no e at the end as Pascale is for women)
[02:43:16] gettys@jabber.org joins the room
[02:43:19] <Lee Howard> merci
[02:43:24] <ray> Jari's presentation is at http://www.ietf.org/proceedings/82/slides/homenet-3.pdf
[02:43:28] <Dave Thaler> I think the arch doc should contain a discussion of this (adding to IPv4 vs rearchitecting IPv4 topolgy too)
[02:43:35] gettys@jabber.org leaves the room
[02:43:53] Satoru Kanno joins the room
[02:44:46] <Ole Troan> fyi: ichiroumakino == Ole Troan (wasn't aware I logged on with gmail account).
[02:45:11] <mcr> It bears remembering that flat IPv4 networks we have today either break when there are loops, or have STP to break the loops. So, it's important to remember that flat IPv4 works only because of layer-2 "routing" protocol.
[02:45:41] <Dave Thaler> and most home nets don't have STP
[02:46:19] <Lee Howard> please, god router, no STP
[02:46:21] <Ole Troan> http://www.ietf.org/proceedings/82/slides/homenet-3.pdf
[02:46:43] <mcr> I've seen it on quite a number of things, just not on by default. Every linux based home gateway has STP code there already, as it's already in the bridge code, and *wrt lets you turn it on.
[02:47:11] <evyncke> zOSPF or how to make OSPF looking like BGP by carrying non routing related information :-O
[02:47:12] <mcr> My precious and I hates the evil STP.
[02:47:50] <Lee Howard> "Requirements" slide
[02:48:37] <Lee Howard> "OSPFv3 Prefix Assignment" slide
[02:49:16] <Lee Howard> "OSPFv3 Distirbuted Prefix Asignment Heuristic"
[02:50:24] lllmartini leaves the room: Replaced by new connection
[02:50:26] lllmartini joins the room
[02:51:25] <mcr> I think I'd be much happier having OSPF distribute a "IPv6-FOO offers DHCPv6 PD" hint in an LSA.
[02:51:55] <Lee Howard> somebody CHakrabarti
[02:52:27] <Lee Howard> ANdrew MacGregor
[02:52:29] <Ole Troan> Andrew McGregor
[02:52:53] <Lee Howard> I cant remember howto spell his name. . . Dec Wej?
[02:53:02] <behcet.sarikaya> Hui Deng
[02:53:16] <Lee Howard> totally different person
[02:53:18] <evyncke> COnfusion with Wojcheck Dec? :-)
[02:53:29] <Lee Howard> yes, theylook exactly alike to me
[02:53:34] <Lee Howard> sorry for confusing the room
[02:53:55] <Lee Howard> Ole Troan I can spell
[02:54:01] <ray> next presentation: http://www.ietf.org/proceedings/82/slides/homenet-4.pdf
[02:54:23] <ray> It was Samita Chakrabati speaking earlier
[02:54:36] <Dave Thaler> except that's not how you really spell Troan :)
[02:54:53] <Lee Howard> it's how *I* spell Troan
[02:55:17] <Ole Troan> Trøan. with a oslash pronounced as the -u in butter.
[02:56:04] <Andrew Sullivan> Gee, and I thought precis was this afternoon. . .
[02:56:13] Warren Harrop leaves the room
[02:56:36] <Lee Howard> did we forget to identify Erik Nordmark at the mike?
[02:56:51] <mcr> I like Erik's idea.
[02:56:54] <Ole Troan> Peter Lothberg at mike
[02:56:54] <Lee Howard> Peter Lothberg
[02:57:11] <Ole Troan> Pete going to propose ISIS
[02:58:17] <Ole Troan> "history / requirements" slide
[02:58:44] <ray> there are no slides for Lee Howard's presentation (due next)
[02:59:09] Ole Troan leaves the room
[03:00:28] Ole Troan joins the room
[03:01:06] angga leaves the room
[03:01:08] behcet.sarikaya leaves the room
[03:01:29] danwing joins the room
[03:02:48] <becarpenter> AH??? I thought that was dead as a dodo.
[03:03:01] <mcr> becarpenter, it's dead for VPNs.
[03:03:10] <evyncke> AH is used for OSPFv3 authentication
[03:03:26] <evyncke> else you are right, everyone uses ESP for IPsec VPN
[03:03:38] <mcr> it's awsome for multicasting stuff out to machines that may or may not have a key to authenticate things.
[03:03:55] <danwing> makes sense, a routing protocol doesn't need to hop over a NAT.
[03:03:58] <mcr> we got scared about it for SEND.
[03:04:12] <Ole Troan> Dave Thaler at the mike
[03:08:11] <mcr> the protocol Erik described supports a RFC6204 at the leaf of the tree.
[03:09:07] <Ole Troan> only if the internal router is pre-configured with what interface is upstream and which ones are down stream
[03:09:37] <Ole Troan> and a 6204 router would fail in the multi-homing case
[03:11:21] TACHIBANA toshio leaves the room
[03:11:24] <mcr> multihoming case, yes, fails.
[03:12:30] <Ole Troan> I do like the leaf-router idea though, but I'm not sure how you would hook that into a zOSPF 'core'
[03:12:33] <mcr> once you add link quality to RA, then you basically wind up with RPL.
[03:13:45] <Ole Troan> upcoming presentation: http://www.ietf.org/proceedings/82/slides/homenet-5.pptx
[03:14:30] Shoichi Sakane joins the room
[03:16:56] <mcr> I think this is fascinating work, but I also think it's out of scope for homenet.
[03:19:35] <ray> you might be right, but hardly anyone else is considering homenet security yet
[03:20:15] <sftcd> I think tihs has a few serious problems as written; the TLS MITM and requiring REPUTE (which is not doing addresses) and a service
[03:22:03] <sftcd> is there going to be any time to comment on this at the mic? (I want to:-)
[03:22:15] <Lee Howard> the service is the part that bother me most. this is a product definition more than a technical spec
[03:22:42] <sftcd> the mitm bit is just mad
[03:22:46] Suz leaves the room
[03:24:06] Satoru Kanno leaves the room
[03:24:47] <Lee Howard> wow, it runs nessus too
[03:24:56] Suz joins the room
[03:25:51] <Dave Thaler> also the allow-by-default model either makes IPv6 less secure (or at least _perceived_ as such) than IPv4, or requires changing the current security expectations (i.e. perceptions) from IPv4. Both I think go against the spirit of "don't break IPv4".
[03:26:53] Lee Howard will abuse the authors privately
[03:26:56] <Dave Thaler> (actually only the second goes against it, the first would give IPv6 a bad name)
[03:27:38] Lee Howard was about to explode
[03:27:52] satoru.matsushima leaves the room
[03:28:36] arifumi leaves the room
[03:29:21] gettys@jabber.org joins the room
[03:29:44] <Ole Troan> http://www.ietf.org/proceedings/82/slides/homenet-2.ppt
[03:29:53] sftcd leaves the room
[03:29:55] Satoru Kanno joins the room
[03:30:06] <ray> not all of those slides will be presented - there's a whole load at the end for extra info
[03:31:06] monden leaves the room
[03:31:27] dudisaki leaves the room
[03:31:58] <Lee Howard> we're up to "Auto Name Prefix"
[03:32:03] <Lee Howard> no, we're past that
[03:32:07] <Dave Thaler> um... the last byte of the MAC address isn't unique
[03:32:16] <Lee Howard> he has a tiebreaker, I think
[03:32:45] ggm joins the room
[03:32:56] <ggm> robert elz wrote a draft on this in the 1990s
[03:33:01] <ggm> does nobody read any more?
[03:33:13] <ggm> mapping v6 into alpha representation alphabet
[03:33:15] <Lee Howard> yes, we read, but we don't read everything ever written
[03:33:23] <mcr> ggm: some participants weren't born yet.
[03:33:28] vikas.sarawat leaves the room
[03:33:31] <ggm> scary, but true
[03:33:36] Satoru Kanno leaves the room
[03:33:38] tomek leaves the room
[03:34:15] arifumi joins the room
[03:34:19] <ggm> (yes, this is different)
[03:34:25] <Andrew Sullivan> I think this work shows that we need to figure out some way of mapping opaque machine-meaningful identifiers to human-useful identifiers. Perhaps we could invent a hierarchical namespace to do this.
[03:34:53] <Dave Thaler> and a protocol that supports reverse resolution
[03:34:54] kawashimam leaves the room
[03:34:59] <ray> Andrew - in part this is a possible solution for the problem we have in V4 of every router having 192.168.1.1 as its default address
[03:34:59] <Suz> Andrew, that trick never works.
[03:35:04] <Lee Howard> it could have distirbuted authority
[03:35:23] <gettys@jabber.org> ray, on my router, it's name is gw.home.lan, not 192.168.1.1....
[03:35:24] <mcr> so, this is a *local* mechanism to map *local* information to something meaningful. It's lets you debug things that seem to have broken your access to the hierarchical and multicasted name space.
[03:35:36] becarpenter leaves the room
[03:35:39] <ray> Jim, yes, and what does the _second_ one get called?
[03:35:53] <ggm> other-gw.home.lan
[03:35:55] <gettys@jabber.org> ray: just -1, -2 and so on...
[03:35:59] <ggm> another-gw.home.lan
[03:36:12] <ggm> this-is-freds-house-he-has-200-routers.gw.home.lan
[03:36:17] becarpenter joins the room
[03:36:17] <ray> which either requires persistence, or user configuration
[03:36:25] <Lee Howard> I like the idea. BUt I'm not sure it's important.
[03:36:44] <gettys@jabber.org> ggm: fred's a special case...
[03:37:00] <mcr> this is useful, and important now that typing "canon" might get me that printer "canon.local", or that site, "canon."
[03:37:16] <mcr> at least, I can refer to it as n7bz-u2.
[03:37:25] lllmartini leaves the room
[03:37:29] lllmartini joins the room
[03:37:36] Olafur leaves the room
[03:37:42] <Andrew Sullivan> we already have llmnr and mdns (and one of these, we even managed to standardize). I don't get the way this is going to solve the inherent problems with those things. Maybe I need to re-re-read the draft, which didn't seem to me to answer any of that.
[03:37:46] <Dave Thaler> wouldn't you rather refer to it as "canon"?
[03:38:06] Colman Ho leaves the room
[03:38:14] ogud joins the room
[03:38:29] <mcr> Andrew, if this does not *replace* mDNS, but rather gives us another handle.
[03:38:36] <Andrew Sullivan> I do fail completely to see how n7bz-u2 is going to be any use at all to Grandma.
[03:38:41] <mcr> Dave, "canon." is their new TLD.
[03:38:48] <ggm> it maps to something meaningful in punycode
[03:38:51] Atarashi Yoshifumi leaves the room
[03:38:53] <Ole Troan> done.
[03:38:55] <ggm> fish-bycicle I think
[03:38:57] <mcr> Andrew: you are completely right. Grandma fails with this name.
[03:38:58] <Lee Howard> it's useful for troubleshooting, not for friendliness
[03:38:59] tomek joins the room
[03:39:03] Ole Troan leaves the room
[03:39:06] tsavo_work@jabber.org/Meebo leaves the room
[03:39:14] ray_atarashi leaves the room
[03:39:14] ggm leaves the room
[03:39:16] h-suzuki leaves the room
[03:39:19] <Lee Howard> bye
[03:39:23] <Andrew Sullivan> bye
[03:39:26] <mcr> it's a useful Informational RFC on what some debian package does.
[03:39:27] tomek leaves the room
[03:39:30] Lee Howard leaves the room
[03:39:31] <mcr> thank you all.
[03:39:31] Chris Griffiths leaves the room
[03:39:50] <Andrew Sullivan> aha! That explains it. I need to be on more Trotskyite lists! Thanks for the clue.
[03:39:58] Andrew Sullivan leaves the room
[03:40:05] <gettys@jabber.org> Lee Howard: good point; can have multiple names, and having something weird for debugging doesn't hurt
[03:40:06] Shoichi Sakane leaves the room
[03:40:44] Suz leaves the room
[03:40:59] naptee leaves the room
[03:41:05] becarpenter leaves the room: offline
[03:41:06] ogud leaves the room
[03:41:17] ray leaves the room
[03:41:24] jweil leaves the room
[03:41:26] yaovct leaves the room
[03:42:00] pselkirk leaves the room
[03:42:18] danwing leaves the room
[03:42:29] babongo leaves the room
[03:42:58] Colman Ho joins the room
[03:44:00] TACHIBANA toshio joins the room
[03:44:10] TACHIBANA toshio leaves the room
[03:44:45] evyncke leaves the room
[03:45:53] lllmartini leaves the room
[03:46:07] ogud joins the room
[03:48:36] tatsuji leaves the room
[03:49:06] naito.kengo@jabber.org leaves the room
[03:51:36] Dave Thaler leaves the room
[03:54:06] ogud leaves the room
[04:01:52] jlcJohn leaves the room
[04:02:06] gettys@jabber.org leaves the room
[04:03:36] arifumi leaves the room
[04:05:25] pselkirk joins the room
[04:06:27] Dave Thaler joins the room
[04:07:50] antoin leaves the room
[04:18:31] lllmartini joins the room
[04:22:50] pselkirk leaves the room
[04:29:39] lllmartini leaves the room
[04:31:54] lllmartini joins the room
[04:37:40] angga joins the room
[04:42:53] lllmartini leaves the room: Replaced by new connection
[04:42:54] lllmartini joins the room
[04:43:09] angga leaves the room
[04:51:07] Colman Ho leaves the room
[04:51:51] lllmartini leaves the room: Replaced by new connection
[04:52:04] lllmartini joins the room
[04:57:00] Shoichi Sakane joins the room
[05:02:12] Colman Ho joins the room
[05:02:48] tomek joins the room
[05:03:30] naptee joins the room
[05:03:47] naptee leaves the room
[05:12:05] Satoru Kanno joins the room
[05:13:15] Satoru Kanno leaves the room
[05:16:37] Shoichi Sakane leaves the room
[05:17:02] Colman Ho leaves the room
[05:17:26] Shoichi Sakane joins the room
[05:24:08] lllmartini leaves the room
[05:24:23] lllmartini joins the room
[05:26:58] lllmartini leaves the room
[05:28:37] Dave Thaler leaves the room
[05:43:49] arifumi joins the room
[05:44:46] naito.kengo@jabber.org joins the room
[06:00:18] Dave Thaler joins the room
[06:02:55] danwing joins the room
[06:08:15] Shoichi Sakane leaves the room
[06:10:08] danwing leaves the room
[06:26:37] tomek leaves the room
[06:28:34] tomek joins the room
[06:33:52] Dave Thaler leaves the room
[06:34:52] tomek leaves the room
[06:35:16] tomek joins the room
[06:35:41] tomek leaves the room
[06:36:02] tomek joins the room
[06:36:25] tomek leaves the room
[07:00:08] arifumi leaves the room
[07:01:54] danwing joins the room
[07:13:35] arifumi joins the room
[07:15:48] naito.kengo@jabber.org_ joins the room
[07:16:21] danwing leaves the room
[07:17:08] naito.kengo@jabber.org leaves the room
[07:18:20] arifumi leaves the room
[07:20:28] Ole Troan joins the room
[07:38:49] danwing joins the room
[07:55:22] Ole Troan leaves the room
[07:57:11] Ole Troan joins the room
[08:29:33] danwing leaves the room
[08:43:16] Ole Troan leaves the room
[08:44:47] Ole Troan joins the room
[09:04:47] Ole Troan leaves the room
[10:09:22] naito.kengo@jabber.org_ leaves the room
[13:22:25] Ole Troan joins the room
[13:47:33] Ole Troan leaves the room
[14:52:46] mcr joins the room
[16:35:46] Ole Troan joins the room
Powered by ejabberd Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!