[16:56:24] stevey joins the room [16:56:28] stevey leaves the room [22:00:00] arifumi joins the room [22:00:08] arifumi leaves the room [22:01:04] Robert joins the room [22:01:35] fujisaki joins the room [22:10:38] ruriham joins the room [22:10:43] arifumi joins the room [22:11:36] lixia joins the room [22:12:26] frodek joins the room [22:12:28] Andrew Sullivan joins the room [22:12:34] Dave Thaler joins the room [22:12:47] oleg.ponomarev joins the room [22:12:54] becarpenter joins the room [22:13:08] rdenisc joins the room [22:13:11] Atarashi Yoshifumi joins the room [22:13:13] Elena joins the room [22:13:21] acting as Jabber scribe [22:13:39] Presenting the Agenda [22:13:48] shep joins the room [22:14:35] nordmark joins the room [22:14:35] fdupont joins the room [22:14:49] Marc Blanchet now presenting the Problem Statement [22:15:02] behcet joins the room [22:15:03] Slide "Related Drafts to PS" [22:15:25] hartmans@jis.mit.edu/owl joins the room [22:15:37] hirocomb12 joins the room [22:16:15] Slide "COntext" [22:16:20] Slide "Assumptions" [22:16:22] Lars joins the room [22:17:08] shamus joins the room [22:17:35] Desire joins the room [22:17:36] bhoeneis joins the room [22:17:50] raj joins the room [22:18:10] pselkirk joins the room [22:19:13] SLide "Symptom" [22:19:22] I think that's _a_ symptom but by no means the only or even the biggest problem [22:20:01] the assumption that you have IP configured is really broken IMHO [22:20:19] fujisaki leaves the room: Replaced by new connection [22:20:23] fujisaki joins the room [22:20:36] in 3G, it's false for economic reasons. In WLAN it's false for battery considerations. [22:20:42] Slide "DNS" [22:20:53] Slide "interface selection" [22:20:59] RalphD joins the room [22:22:26] btw, slide set is at http://www3.ietf.org/proceedings/09mar/slides/mif-1.pdf [22:22:42] slide "address selction" [22:22:51] (10) [22:23:34] Yeah, but considering the custom of changing slidesets at the last minute, I wouldn't trust the page number on Jabber... [22:23:41] slide Today [22:24:09] Next presentation by Margaret Wasserman [22:24:15] http://www3.ietf.org/proceedings/09mar/slides/mif-2.pdf [22:24:22] slide "Goals..." [22:24:37] good point... [22:25:12] slide "systems currently included" [22:25:41] slide "2 General approaches" [22:27:47] slide "split DNS" [22:30:01] slide "interface selection" [22:30:10] matthijs joins the room [22:31:07] Elena leaves the room: Replaced by new connection [22:31:07] Elena joins the room [22:31:43] slide "iface sel Default Rtr" [22:34:49] slide "addr sel" [22:35:24] Linux does this too. It's been a mess. [22:36:36] all proper getaddrinfo() implementation do [22:36:45] As regards Linux, glibc does [22:36:51] uclibc and dietlibc don'tc [22:36:53] as does Windows getaddrinfo() [22:37:43] Dave Thaler: I think hartmans "too" referred to Windows :) [22:37:56] ah yes [22:38:10] rfc5206 End-Host Mobility and Multihoming with the Host Identity Protocol [22:38:11] Lars leaves the room [22:38:14] slide "going fwd" [22:40:55] next presentation by Gabriel Montenegro [22:41:05] http://www3.ietf.org/proceedings/09mar/slides/mif-3.pdf [22:41:20] slide "How is MIF != from other efforts" [22:41:21] raj leaves the room [22:41:43] raj joins the room [22:42:05] slide "default addr sel" [22:44:21] slide "handling of multiple addresses or ifaces" [22:45:35] raj leaves the room: Replaced by new connection [22:45:35] raj joins the room [22:45:36] HIP assumes that the other sides implements HIP [22:45:44] from implementor PoV, this is a big NO NO [22:46:50] yes, I personally would like to have routable HITs, but many won't agree [22:47:05] same pb with SCTP. I don't know SHIM6. As for MIP, it requires network collaboration [22:47:37] next slide, "PMIP and addr sharing across ifaces" [22:48:20] it would be very useful to know more how to use multiple interfaces [22:48:45] that was explained in the previous presentation [22:48:49] with HIP we did some experiments, but it is not at production level [22:48:59] typically, the IP stack is silo-ized [22:49:12] and apps need to "somehow" select the interface they're using [22:49:16] rdenisc: sure, I just want to say that I am interested in any result :) [22:49:35] Margaret said we have 11 minutes left for the whole BOF, but BOF is scheduled to end at 16:10 (twenty more minutes) [22:49:43] just buy an S60 smartphone, or whatever [22:49:45] with HIP, we just try to select interface on behalf of the applications and the same (policies etc) applies [22:50:28] that's not workable. you need to select the interface before you can even query DNS due to split horizons [22:50:53] application gets HIT from DNS [22:50:59] this is very prevalent with VPN and semi-isolated corporate networks [22:51:06] if the domain has HIP record [22:51:50] but that's not backward compatible with the existing internet. [22:52:06] what implementors (incl. me) want is something that works today [22:52:31] sure [22:53:31] that is why I want routable HITs, so HIT can be the source IPv6 address in routed packets [22:54:12] raj leaves the room: Replaced by new connection [22:54:12] raj joins the room [22:54:15] one interface per mic... oops we have an interface selection problem :) [22:56:17] matthijs leaves the room [22:56:18] lixia leaves the room [22:59:15] lebobits joins the room [22:59:26] raj leaves the room: Replaced by new connection [22:59:26] raj joins the room [23:01:55] kanda joins the room [23:02:43] jjones joins the room [23:03:20] pselkirk leaves the room [23:03:29] oleg.ponomarev-- you say you want routable HITs. Sounds like you want SHIM6 instead. (I don't think HITs should be routable, but the addresses used in SHIM6 do have some (most?) of the things you would want from a HIT, but are also routable (normal) IPv6 addresses. [23:04:04] [and shim6 addresses can be secured] [23:04:35] ... and internally, SHIM6 and HIP are pretty much the same thing (if you take the cover off and look at what an implementation looks like). ) [23:04:35] but it ends with '6' [23:05:11] (It makes good sense to implement SHIM6 and HIP together in a unified way.) [23:06:30] Atarashi Yoshifumi leaves the room [23:06:34] becarpenter leaves the room [23:06:34] thanks everyone for bearing with me [23:06:37] shep: Yes, might be. I should check details of shim6 [23:06:43] Andrew Sullivan leaves the room [23:06:50] rdenisc leaves the room [23:06:53] oleg.ponomarev leaves the room [23:06:55] behcet leaves the room [23:06:57] ruriham leaves the room [23:06:59] Elena leaves the room [23:07:13] jjones leaves the room [23:07:17] nordmark leaves the room [23:07:50] fdupont leaves the room: Computer went to sleep [23:08:10] shep leaves the room [23:08:11] raj leaves the room [23:08:47] kanda leaves the room [23:09:22] shamus leaves the room [23:10:23] arifumi leaves the room [23:10:55] frodek leaves the room [23:12:16] lebobits leaves the room [23:15:55] Robert leaves the room [23:17:54] kanda joins the room [23:20:07] hirocomb12 leaves the room [23:22:55] RalphD leaves the room [23:23:54] bhoeneis leaves the room [23:25:04] Dave Thaler leaves the room [23:25:20] fujisaki leaves the room [23:53:08] Dave Thaler joins the room [23:54:21] Dave Thaler leaves the room