[21:57:39] agallant joins the room [21:58:45] agallant leaves the room [22:09:36] Dan Burnett joins the room [22:10:14] Adam Uzelac joins the room [22:10:36] not_a_scribe joins the room [22:12:03] bernie joins the room [22:12:06] bpenfield joins the room [22:12:52] ray joins the room [22:14:18] pfaltstr joins the room [22:14:36] Meeting is called to order by people in the audience. The chairs woke up. [22:15:11] Chairs present themselves. [22:15:17] bernie leaves the room [22:15:40] ray leaves the room [22:16:10] Cullen Jennings joins the room [22:16:55] There are problems with the projector, so the meeting has not really started yet. [22:17:29] Cullen Jennings got a token of appreciation by the chairs. [22:17:32] as0-d91k joins the room [22:17:40] dcrocker joins the room [22:17:41] agallant joins the room [22:19:03] POLL: say "aye" if you are a remote participant [22:19:13] aye [22:19:20] Presentation of the background by chairs. [22:19:33] @dcrocker - ACK [22:19:53] Problem looks at mapping from E.164 to metadata, which can not be ENUM as ENUM returns explicitly a URI. [22:20:01] (150 feet, but still...) [22:20:09] There is a draft protocol, use cases, problems, draft charter etc. [22:20:40] Hyong-Jong Paik joins the room [22:20:55] http://www.ietf.org/proceedings/10mar/slides/e2md-2.pdf [22:21:05] BCP44 always applies. [22:21:24] Discussion on whether this protocol is used on global Internet or just local. [22:21:31] Bernarda joins the room [22:21:35] Cullen explained the importance is to explain the use case. [22:21:47] Bernarda has set the subject to: E2MD BOF [22:22:14] suzukisn joins the room [22:22:17] BCP 44?? "Use of HTTP State Management" [22:22:18] Klensin: Maybe you should not call it "ENUM" if it is not for the public Internet. [22:22:34] dcrocker: that was the word that was used at the mike. [22:23:00] ack. tnx. [22:23:01] [presentation of slides on the URL provided earlier] [22:23:05] can anyone see if the audio stream is working [22:23:13] hard to tell if mics in room are working [22:23:35] audio works [22:23:46] "ENUM indicates a resource" [22:23:56] dwillis joins the room [22:23:57] thanks - on checking audio recording [22:23:57] slide 4/12 [22:24:35] slide 5/12 [22:26:11] more slides.....now 8/12 [22:26:54] comment on send-n: has been used in the UK to implement number portability database, indicate how far down the tree you have to go to find terminate nodes in the tree [22:27:02] slide 9/12 explains [22:27:10] HKAPLAN1 joins the room [22:28:13] send-n give information about how many more digits are needed, as you normally only know "more digits are needed" (but not how many) [22:28:25] Slide 10/12 [22:28:33] Andrew Sullivan joins the room [22:29:53] Goal with cnam is to ensure we can pass around the data that exists in the PSTN in SIP -- using IETF architecture. [22:30:05] slide 11/12 [22:30:10] naptee joins the room [22:31:23] John ask: What is metadata? [22:31:59] Q: The three use cases are so different that the question is whether there is any commonality part from "they are not ENUM". [22:32:01] trit's really _data_ [22:32:32] A: They are similar and more common than what one think, part from c.* that is only looked up at the end node. [22:32:36] it's data, not meta-data [22:33:17] Klensin: other things you describe are things we do not try to store in DNS. Note that we do not store mailboxes or web pages in DNS. We store references to them. [22:34:51] bpenfield leaves the room [22:35:38] long line with the queue and I can not type all things [22:36:08] Dean Willis: The question you asked John is still an open issue. [22:36:48] Dean Willis: If we need some storage that is similar to DNS, we can as well use DNS. [22:36:59] Peter Koch: yes, cnam is different [22:37:48] bpenfield joins the room [22:38:14] Peter Koch: Anything that tries to cut the tree, jump the tree etc are not specific to the ENUM or E.164 numbers, are general to DNS. [22:38:27] The argument that because you have to get to _mostly_ the same functionality as the DNS, you might as well use the DNS, fails in case the additional stuff you need is a bad fit in the DNS [22:39:04] Olaf: DNS is designed for query/response, so ENUM matches fairly well. [22:39:57] Olaf: However, if one go down the path of putting metadata in the system, the question is what kind of things you need for finding the metadata. Is a query/response system enough, or do you need search? If you need search, then you might need a record that reference the service where the metadata is stored. [22:40:02] bernie joins the room [22:40:18] Dean Willis: Open issues [22:40:46] http://www.ietf.org/proceedings/10mar/slides/e2md-3.pdf [22:41:19] Alan Johnston joins the room [22:41:20] ajs: mikrophone line was closed, but I will remember what you wrote, and say it when it "works" [22:41:44] I'm here. I put it here because the mic line was closed :) [22:41:48] :-) [22:42:02] ogud joins the room [22:43:23] Klensin: Question is whether ENUM do work well enough to "just inherit" the design of the owner in DNS, or rather whether the statement that it is "optimal" is correct. [22:44:07] John: ENUM have from the beginning had problems with security issues, disclosure etc. [22:45:00] Maybe we should just copy the DNS RFCs, do a s/DDDS/EDDDS/ and s/DNS/ENS/ or whatever, and get away from dealing with "DNS" once and for all [22:45:17] oh and change from port 53 to some new port [22:45:26] Slide 3/6 [22:45:38] Andrew Sullivan leaves the room [22:45:39] Cullen Jennings leaves the room [22:46:55] Cullen Jennings joins the room [22:47:34] Yes, if this is of good use in private networks perhaps an equivalent alternative intended for such networks should be created. Strong use means there is value in such a model (even if it should be a duplicate of DNS rather than DNS itself). [22:47:35] Cullen: If having a ringtone in DNS is not a good idea, what things are good ideas, and how do we know? [22:48:19] Slide 4/6 [22:48:46] Cullen Jennings leaves the room [22:48:47] Cnam also happens to be a name resolution if you think about it - it's just resolving a E.164 to a string name [22:50:39] slide 5/6 [22:52:04] slide 6/6 [22:53:38] paf: mutiple problems with NAPTR (written up by IAB).... [22:54:01] paf: Should not have used an NAPTR RR when using ENUM..... [22:54:11] paf: Should start by writing down what you need.... [22:54:31] paf: Size of the records, how you are finding them, what kind of RRs match the requirements... [22:54:41] paf: Should you create another RR? [22:54:50] Jon: Would one RR fit all? [22:55:01] Thanks Bernard. [22:55:23] can someone go to the mike: does using NAPTR for cnam cause harm?? [22:55:48] will ask! [22:56:57] Dean: SRV RRs use a different syntax... another model is to use _cnam.... [22:57:10] Dean: would require defining multiple RRs.... [22:57:27] Dean: wouldn't require defining mutiple RRs (correction) [22:57:47] Pen joins the room [22:58:07] but we use EDNS0 and are proposing to have a nuique key using "_cnam.1.2.3.4.5.5.e164.arpa" [22:58:13] patrik: size of response from query might grow. [22:58:14] ray joins the room [22:58:39] that last point about multiple records for the same (QNAME, QTYPE, QCLASS) is exactly the same in 3761 ENUM [22:58:52] hadriel: yes, no comments have been made on the actual proposals, discussion is about whether the requirements are listed detailed enough. [22:58:55] jaap joins the room [22:59:23] dave: if we are after a religious debate, we can have that... [22:59:34] dave crocker: what problems are resolved? [22:59:41] dave crocker: what are the specific queries? [23:00:13] Barnie: use cases have been proposed to the list, 10 extra added lately [23:00:47] Ray Willis: we do it here as the ENUM wg refused to take up it as a work item [23:00:54] nice [23:00:55] cullen: fair enough [23:01:08] cullen: what would be the registration procedure [23:01:41] cullen: are we here because the registration procedure is broken in other documents? [23:01:48] dcrocker leaves the room [23:02:02] rich: the intent is trying to follow the existing enum registration procedure. [23:02:52] Dave Schwarts(sp?): I think NAPTR is inadeqvate is because the data is about routing or information about higher levels. [23:03:18] unused helps with a routing decision [23:03:38] send-n is a routing decision too, now that i think of it [23:03:40] q from chair: is somewhere under this pile of stuff some need in IETF to work on this? [23:03:53] a: "a lot of people" [23:04:31] q: do you think dns is the right place to start this? [23:04:33] PAF - It's "Bellis" - and the problem was Layer 9+ - they weren't refused for technical reasons [23:04:42] dave: I do not understand the order of the questions. [23:04:44] Ray: thanks [23:05:02] dave: you do not have specific problems you try to solve [23:05:18] dave: you are asking for responses to what feelings people have. [23:05:28] q: is the cnam use case good for DNS? [23:06:20] I put my virtual hand up [23:06:54] can someone put my hand up on the mic? :) [23:06:54] for which vote? [23:06:59] for DNS of course [23:07:23] answer: 10 in favor, 2-3 against (I do not see exact against) [23:08:27] Christer Holmberg: is not cnam something you already can do with sip option? [23:08:40] answer: yes, potentially [23:09:04] sure, 20 seconds later the phone rings... great... [23:09:41] q: how many people feel they understand the problem scope [23:09:49] a: ~32 [23:09:53] YES [23:10:01] 33 [23:10:02] q: How many do not think the problem scope is clear enough? [23:10:11] a: "some" [23:10:18] ogud leaves the room [23:10:36] q: are there people here that want to solve the problem [23:10:38] YES [23:10:39] a: a fair amount [23:10:46] (but not in the room) [23:10:52] Adam Uzelac leaves the room [23:10:58] bpenfield leaves the room [23:11:03] ray leaves the room [23:11:11] q: who read the charter? [23:11:20] q: can we build a wg on that charter? [23:11:29] not_a_scribe leaves the room [23:11:36] YES [23:11:37] a: 5-8 rised their hands [23:12:14] q: is the current charter not ready to go [23:12:21] a: a larger number of hands [23:12:22] Dan Burnett leaves the room [23:12:27] meeting adjurned [23:12:31] naptee leaves the room [23:12:33] Alan Johnston leaves the room [23:12:38] pfaltstr leaves the room [23:12:48] as0-d91k leaves the room [23:12:48] suzukisn leaves the room [23:12:54] HKAPLAN1 leaves the room [23:15:12] agallant leaves the room [23:15:54] Pen leaves the room [23:23:18] dwillis leaves the room [23:25:18] Bernarda leaves the room [23:30:47] jaap leaves the room [23:31:38] bernie leaves the room [23:36:00] Hyong-Jong Paik leaves the room [23:36:32] Bernarda joins the room [23:38:21] Bernarda leaves the room [23:42:28] ogud joins the room [23:44:52] ogud leaves the room [23:53:06] Andrew Sullivan joins the room [23:53:25] Andrew Sullivan leaves the room [23:54:10] ray joins the room [23:54:13] ray leaves the room