RIPE 55 ENUM WG Minutes
Thursday, 25 October - Amsterdam
Chair: Carsten Schiefner
Scribe: Alex Le Heux
Webcast and Feedback Archives:
A. Administrative Matters
- Select a scribe
- Jabber Monitor
- Microphone Etiquette
- Finalise agenda
Alex Mayhofer (enum.at) would like to present about crawler.enum.at website if there is time.
B. Minutes from Working Group Session at RIPE 54
Carsten: The previous minutes were finalised on the mailing list, there was only one typo, which the chairs will take care of.
C. Review Action List
54.2 [Carsten Schiefner] - with RIPE NCC Services about DNSMON: -> CLOSED
54.3 [Niall O'Reilly]: Communicate result to WG Chairs, RIPE-NCC
[Niall O'Reilly] Discover from group whether ENUM WG session could be held in parallel with AP WG and feed back to WG-Chairs and RIPE NCC. No objections followed from e-mail on this to the WG list. None were made at the meeting either. Niall will communicate this, and this option may be used in planning future meetings.
OPEN Expected to be closed at WG-Chairs meeting immediately following WG session.
ENUM-AP-52.2 [Carsten Schiefner]: NCC correspondence archive: -> CLOSED
Carsten proposed to close this. There were talks and the RIPE NCC agreed to put back the archive as it used to be, including all correspondence, including e-mails and scanned letters etc. The RIPE NCC is already working on this.
ACTION 54.1: Think about the task and scope of quality checks and form a new Quality TF. Take this to the mailing list: -> CLOSED
Carsten proposed to close this. What was meant to go on here is in line with the DNSMON effort. There were no objections.
D. Main Presentations
D1: Motivating ENUM: Avoiding the Bottlenecks - K. Nieminen
There were no questions.
D2: Infrastructure ENUM - Implementation Options for +31 - L. Maris/P. Nooren
Carsten: Do you have plans to extend this research into other countries, especially those with no centralised number portability database, like Germany?
Pieter: No, we see a move from having no central DB to having a central DB using ENUM, like in the UK.
D3: ENUM: a heretic's view on SIP routing - O.Lendl
Geoff Huston: Firstly, you seem to have a denial about the PSTN. The phone system works off bilats across the globe, there is no transit. The phone system can't handle three party interactions. It's all bilats. Because you don't know how to divide the money between three parties, let alone fifty. But the physical world isn't quite like that. If I haven't got enough capacity and you do, we do a brokerage deal. Then you sort of say we can't do bilats, so somehow we need to do this policy routing thing. The keyword is "overlay". Two people can do the deal with payed settlement, across IP, because IP pushes packets. I can still do the VOIP settlement and the push in and out of the PSTN. You seem to go from "bilats are impossible for the world" to "I need policy routing" to "but I need this really strange policy routing, which is not per policy, or per application, but per indication per application".
Otmar Lendl: There is a lot of transit in the PSTN. Like a small Austrian provider has a call to Congo, it passes it to Belgacom or so who pass it on. This routing info needs to filter back.
Geoff: No. inside your international switches are entries for almost every operator, because we do not know how to split the money, we can't do transits. It's weird, but it works.
Jim Reid: Agrees with what Geoff says. The way we try to solve these problems, like with Spearmint, is in my opinion the wrong direction. We try to find something which is more culturally compatible with the way telcos and carriers do business amongst themselves, which is, like Geoff said, by bilateral agreements. It may well be that these agreements are reflected in the DNS. That might be a beter way forward than Spearmint. I don't like the idea personally of peering federations, balkanised trees. that's just madness. We're just creating more problems and more fragmentation and more diverse numbers of namespaces. We need to get back to the idea of single tree, single namespace. This is what we or the IETF should try to fix.
Otmar: What happens is two operators don't share a tree?
Jim: Same way they do it now, with steamphones. (Those without internet addresses, with keypads and rotary dials, steam powered)
Patrick Faltstrom: You compare with e-mail, that's interesting, and come to conclusion that the work to stop spam should be done on the sender side, I agree. I don't agree with your conclusion that the end-to-end model is dead. The IETF had the discussion if we should move from e-mail today to consent based messging. But the ability to make a call or e-mail to someone without prior consent is worth a lot. So the the Spearmint and voice people can probably learn a lot more from the e-mail people.
Vladislav Vobr (Emirates Telecomunication Corp.): Why do you still need ENUM if PSTN is gone?
Otmar: Because we still have numbers.
Vladislav: I see ENUM as temporary. Do you have an exit strategy for ENUM, you think it's going to stay forever?
Otmar: No, numbers are the least common denominator for identifying a voice endpoint, they work everywhere.
Vladislav: we moved from IPs to names, much easier for people to understand, it's the same for the phone.
Otmar: Are you sure you can understand a Chinese name better than a number? How do you enter that?
Vladislav: That's the next step: IDNs, international domain names.
Otmar: The amount of deployed phone numbers is so huge, that it's very hard to replace.
Vladislav: I think it will.
F. Short News
This maybe curtailed in case of time pressure.
F1: Plan for Signing of E164.ARPA - A.Robachevsky
There were no questions.
F2: +44 update - J.Reid
There were no questions.
F3: +353 update - C.Daly done by Niall, as C.Daly can't travel here.
People were asked to pass on questions to the presenter, as Niall gave the presentation on behalf of someone else.
G: Discussion on Plenary Presentation: Emergency Call Handling - A. Mayrhofer
Carsten asked if any discussion for this needed. No one responded.
X: Interaction With Other Working Groups
X1: Extension of DNSMON to E164.ARPA (NCC Services & Test-Traffic WGs)
Carsten: This is about the interaction with the RIPE NCC Services and Test Traffic WGs and the DNSMON service being extended into the ENUM name space as well. We had a presentation by Ondrej Sury in the RIPE NCC Services WG meeting. If you have any questions or comments, please see the main discussion thread in the RIPE NSS Service WG mailing list. This is mainly a heads up that the discussion will be held there.
Jim: I've got one thing that I'd like to emphasise: Everyone knows where I stand on DNSMON in general, but I've lost the argument. On the other hand, if we proceed with this for tier-1 delegations, we need some clarity about how this is going be monitored and get some form of consent from the people who are going to get the monitoring service on how this is done. It would be unwise if this goes ahead and we have the situation where a country perceives that the RIPE NCC is intervening in what is really a national matter.
Carsten: You propose that this is then written down in the policy document?
Jim: Yes! Then if there are complaints, we can show them.
Wilfried Woeber: Do we have time to restart this discussion here or should we move this to mailing list? If they offer a service which by purpose and architecture is to be used by the general public, then why worry about using that service publicly? It makes no difference what kind of query you send. If you do not want it to be public, don't offer it.
Peter Koch: I appreciate this effort to make proposal. How many tier-1 registries are we doing this for now that are not monitored now? Isn't this discussion in vacuum?
Carsten: I can't answer this. at the end of the day it needs to be discussed by greater public.
Niall O'Reilly: This will be taken to the RIPE NCC Services mailing list for input.
Carsten: I'd like to mention the crawler presentation and maybe postpone it. It was decided that this would be moved to RIPE 56.
Summary of action items
- ENUM54.3 [Niall O'Reilly]: Communicate result to WG Chairs, RIPE-NCC
- ENUM55.1 [Niall O'Reilly, Carsten Schiefner]: Update the RIPE54 minutes to take out the typos.
- ENUM55.2 [Niall O'Reilly, Carsten Schiefner]:Take account of offer crawler presentation in planning WG session for RIPE 56