OK, Thanks Rui/peter/james/alex/victor/all, I can see we're talking at cross purposes. The main thing, throughout all of this enum stuff, is to encourage/force some interoperability between various vendor's (let's call them) 'unified messaging' solutions. If that's not the primary problem/challenge, could someone tell me what is? In the first instance, we should have no interest in the technical considerations. We should just look at the potential market and figure out how, if we we're able to come up with an approach which pulled a market, any private vendor would want to jump on it. So let me illustrate the potential market. I.e. global groups. To be clear. "Groups" tend to fall into these three divisions. https://www.ripe.net/ripe/groups We can see that there have been many discussions about ENUM going on around various domains for many years. E.g. http://www.ripe.net/ripe/mail/archives/enum-wg/ So that's one indication of one global conversation, around some piece of Research, which takes place, in different languages, in hundreds of domains, at any one time. Even when these researchers get together at conferences, their records are buried around various institutional and association web sites. We all know that because, since the web was invented, we can all see it. In the case of RIPE, we will discover what its institutional peers have in common only by reading documents which are buried on some other (higher order) domain. E.g. http://www.nro.net/wp-content/uploads/RIR_WTPF13_2.pdf This is not an unusual situation. Everyone within some sphere of practice is siloed, and often try and overcome their limitation by inventing another institution, domain, & web site. And keeping the same discussions separate. While this fragmentation of effort is now obvious, we have librarians who, because the limit themselves to manhandling information/papers/journals (and ignore Communications solutions as studiously as Communications technicians ignore their needs) must pay publishers to aggregate journals on behalf of global (peer) groups. Bringing it down to our present situation. Our BOF, at this terena domain, have our enum discussion, which will have no record unless the elist has an archive. So while James will suggest the OVCC guys have an interest in our discussion, and I point to the Ripe enum group, and Peter I am sure will have some interested parties inside Geant and Dante, there is simply nowhere in domain in cyberspace we might all look to as a spot where other communities with an interest in ENUM might reside, or have resided. The demands for collaboration are obvious, and we all have ideas about real time & asynchronous tools, But that's not important until we have some means of finding our global peers, or peers who might have proceeded us in the same inquiry. If you consider a solution to this problem, You'll be thinking (perhaps, in terms of a new TLD, say) .research. Or using an existing domain like .edu. Regardless, if you are a librarian it will be quite an obvious thing to consider using an old classification scheme for the domain. What is, using the existing way of classifying domain names called http://ddc.typepad.com/ is just as recognizable to a (dewey) librarian as www.025.431.com Now the only shortfall of this approach, using six numbers, is that it can only define 999,999 virtual domains. But it's strength is that these become not just pointers to a million transitional domains, but long term archives in their own right. It's the policies, which are made for the use of virtual rooms - interoperable, open source, open access, IPv6, etc - where the real influence resides. A Global user group doesn't get issued one unless they agree. So you might understand. I don't see the domain being issued by network, or service, operators. They get issued by librarians. If we get that far, we'll need a prefix to be tacked on to a room's number, so people can dial in to a room/conference, using the PSTN, without incurring toll charges. I hope that scopes the concept clearly. All the best, si -----Original Message----- From: Rui Ribeiro [mailto:rui.ribeiro@fccn.pt] Sent: Wednesday, 29 May 2013 6:14 PM To: 'Simon Fenton-Jones'; 'James Sankar' Cc: discussion@nrenum.net; tf-media-prep1@terena.org; 'Alex Galhano Robertson'; 'Victor Reijs (work)'; 'Peter Szegedi'; enum-wg@ripe.net; 'Juan Quemada' Subject: RE: [discussion] Re: Global Virtual Numbering Hi all, {{ Something like this would be possible: 00 ### @ %%%%%%%% 00 ### @@ %%%%%%% 00 ### @@@ %%%%%% 00 ### @@@@ %%%%% 00 ### @@@@@ %%%% Can't figure out why we'd need a network number though. }} I see two reasons upfront: 1. TERENA (or NRENum.net admins) don't have the structure to delegate directly every block. 2. Organizations (NRENs, PCMG's*, ...) will want to delegate some blocks on their own or "transpose" their current numbering into the "global tree". * PCMG's = entities like: Polycom, Cisco, Microsoft, Google, ... Rui Ribeiro Gestor Serviço Técnico de Vídeo GEANT3 - SA3 - T4, eduCONF Leader FCCN Aviso de Confidencialidade/Disclaimer Esta mensagem é exclusivamente destinada ao seu destinatário, podendo conter informação CONFIDENCIAL, cuja divulgação está expressamente vedada nos termos da lei. Caso tenha recepcionado indevidamente esta mensagem, solicitamos-lhe que nos comunique esse mesmo facto por esta via ou para o telefone +351 218440100 devendo apagar o seu conteúdo de imediato. This message is intended exclusively for its addressee. It may contain CONFIDENTIAL information protected by law. If this message has been received by error, please notify us via e-mail or by telephone +351 218440100 and delete it immediately -----Original Message----- From: James Sankar [mailto:James.Sankar@aarnet.edu.au] Sent: Tuesday, 28 May 2013 8:47 PM To: Rui Ribeiro; 'Alex Galhano Robertson' Cc: discussion@nrenum.net; tf-media-prep1@terena.org; 'Victor Reijs (work)'; 'Simon Fenton-Jones'; 'Peter Szegedi' Subject: Re: [discussion] Re: Global Virtual Numbering Dear All I know that this issue is a common one in the commercial world and has the added complexity of providing revenues to justify investment. The Open Visual Communications Consortium (OVCC) are working with carriers, managed service providers, video vendors and supporting service providers to do just that. Given we seem to be redefining the objectives and goal it may make sense to gain an insight into the issues and challenges that the OVCC face so that we can take that effort and any learnings into account. If this would be of value I can organise OVCC leaders in the technical WG to meet via video. Best wishes James Sankar On 28/05/13 8:27 PM, "Rui Ribeiro" <rui.ribeiro@fccn.pt> wrote:
Hi Alex,
{{ I insist, we need to clearly define our objectives with this global numbering plan. }}
IMHO, we need a way to allow addressing of IP voice and video terminals within the global numbering plan that is widely used, without the restrictions of the current numbering plan.
Restrictions = number associated with phone service.
{{ - We want to integrate videoconference and telephony services. }}
True.
{{ - We want a number-only global addressing scheme. No alfabetic characters. }}
"In the future", URI dialing will be the way. I think that we all acknowledge that. But, for the moment, numbers are needed and the correct approach to "route" numbers into URIs is using ENUM. Yes, I would look for a "number-only global addressing scheme".
{{ - We want it to be "E.164 compilant", so we could call from any videophone or even any standard phone. }}
There is a subtle thing that is mixed on this sentence. First, yes, we should be E.164 compliant, as this would promote integration with the current PSTN network and ease future integration. The other thing is to make a call... I would like, but it isn't a MUST today, to be able to dial from a (pstn) phone to these "number-only global addressing scheme". "In the future", if we are an extension of e.164, then it would be easy for operators to route calls from their terminals/phones.
{{ - The translation to URI will be done by NRENum.net. }}
Yes.
{{ 0- Do we really need virtual numbers? IMHO, we dont need it. At least, the majoroty of our NRENs does not need it. But we want it because some countries (its universities and even NRENs) have dificulties to obtain real telephone numbers and make this real numbers the addres of its videoconference rooms. }}
True. Look at the wider picture and think of other service providers that, already promote V&V services within their own numbering plan. What if there was an global service that would accommodate these numbers/networks?
{{ As the answer is "we want it", let´s go on with other questions... }}
Ok. ;-)
{{ 1- What we should address? A- people: our office deskphone or our personal videoconference software B- groups: group of deskphones C- video resources: MCU rooms, physical room "CODECs", Webconf rooms (if we can make it interoperate, etc) D- all above
I think letter C is the correct one, because we already have one telephone number on our business cards. And it is easy to map those addresses on NRENum.net. It is already done for most of us, I think. }}
Option "E" - All of the above and whatever comes up next! Numbers are just numbers... they will then be used for whatever purpose people wants to use them. We should not block their usage upfront by saying "only to be used by physical terminals". There should be rules about the quality of the records, but other than that, we should promote innovation of services.
{{ 3- How should the scheme be? A- Hierarchical and country code dependent? B- Flat, non-hierarchical, independent on any existing coutry code
I think this is the key question to be answered! It would be much easier if we choose leter A. It is already done for most of us. But some people prefer the first choice: flat and CC independent. }}
If A is chosen, there will be three problems: 1. (must!) exist country organization to take care of the service with all the problems associated with it: uptake, different rules, business case, politics, ... 2. bigger numbers (at least more 3 digits...) 3. doesn't solve supra national entities/networks.
I'm in favor of B, but based on the concept of networks. There should be a "world wide registry" for these networks and then there will be registrys to take part of those numbers. Of course NRENs could apply to take care of the CC code, if they want to, but this would open up the management outside country boundaries. For instance, on eduCONF, the numbering that we aim is supra national, where should it be placed on a CC code hierarchy. Where should, for instance, Polycom/Cisco/Microsoft/Google register costumers?
(please don't flame me on these reference to companies (PCMG). They are a big part of the videoconference ecosystem, preparing the model to allow them to enter is, IMHO, a wise decision)
{{ If we decide for option B, I would like to make some other questions about that IPv4 aproach I proposed a few days ago. }}
We could use a "network" prefix to that kind of numbering... just like any other network (educonf, country's nren, PCMG, ...) it would be interesting because it could be registration free and hassle free. In fact most of it would be able to be automatically assigned. But it isn't a "silver bullet" to solve the issue because many terminals are behind nat and you can't address all the terminals these way.
Again... the wide/open the model, the more distinct and innovative the options will be.
{{ 1- How many NRENs or Universities are using NATed IP addresses in its MCUs? As H.323 and SIP is not so good with NAT, I dare say no NAT and no private IP addresses (10/8, 172.16/12 or 192.168/16) are been used for these video resources. }}
Many... most (if not all!) of our VoIP network is behind an SBC that is the only "public" IP address known on the open network. H.323 is even less good transversing NAT, nevertheless there are many using that architecture.
{{ 2- How many of these video resources are using IPv6? Who has today an MCU using just IPv6? When do you plan to remove IPv4 from these devices? }}
There are IPv6 resources available, all Cisco (Cisco, Tandberg, Codian) products are IPv6 enabled. Others like Radvision, Lifesize and Polycom are taking major steps. There are some NRENs that are providing the service in IPv6.
Of course there isn't a date to remove IPv4. The question, is when will it be impossible to add terminals to the IPv4 network. When it happens, we need to be prepared.
As final comment:
The structure that I propose is something like:
00{INP}{NP}{NUMBER}
INP = Internet Number Prefix CC (as "883", or something similar...) NP = Network Prefix to be assigned by the global registry. It could be from 1 digit to 5 digits (?). Where NP that are equal to E.164 CC could be reserved to NRENs or regulators of those countries. NUMBER = extension number within the network addressed by {NP}. The total number should not be longer than 14(?) digits.
Something like this would be possible:
00 ### @ %%%%%%%% 00 ### @@ %%%%%%% 00 ### @@@ %%%%%% 00 ### @@@@ %%%%% 00 ### @@@@@ %%%%
Eventually the INP should be just 1 digit, but that would send us to another "numbering war". ;-)
This structure would allow us to delegate about 40k networks...
Rui
Rui Ribeiro Gestor Serviço Técnico de Vídeo GEANT3 - SA3 - T4, eduCONF Leader FCCN
Aviso de Confidencialidade/Disclaimer Esta mensagem é exclusivamente destinada ao seu destinatário, podendo conter informação CONFIDENCIAL, cuja divulgação está expressamente vedada nos termos da lei. Caso tenha recepcionado indevidamente esta mensagem, solicitamos-lhe que nos comunique esse mesmo facto por esta via ou para o telefone +351 218440100 devendo apagar o seu conteúdo de imediato. This message is intended exclusively for its addressee. It may contain CONFIDENTIAL information protected by law. If this message has been received by error, please notify us via e-mail or by telephone +351 218440100 and delete it immediately