: [enum-wg] COCOM & ENUM ...

 
Marco writes:
> 
> Very, very unlikely  we have a single tree for operator ENUM. 
> Operators will never be able to agree. And perhaps it  is 
> right - different trees may serve different applications, 
> different operators' communities with different policies.

This may work for some time and if there is a PSTN for default routing
But what if the PSTN is gone?
Your customers will expect that they may call all numbers, so a provider
needs to interconnect with all others (some way or the other).

> 
> Not sure how we can enforce across different 
> communities/trees  the rule that  an operator can only insert 
> his numbers - A job for ITU-T?

I do not think this is necessary. In crime one of the basic questions
is the motivation.

What is the motivation for a provider to enter numbers he is not
terminating,
if you do not get termination charges?

What is the meaning of transit on the Internet?
- Transit of signalling messages = proxying
- Transit of media stream? If the meadia stream does not even touch
the origination and terminating proxy?

It simply does not make sense.

What makes sense if you have more than one tree may be a clearing house
proxy
with access to multiple trees, but they will not get charged per call
minute.

-richard


> 
> marco
> ----- Original Message -----
> From: "Stastny Richard" <Richard.Stastny at oefeg.at>
> To: "Jim Reid" <jim at rfc1035.com>; "Richard Shockey" 
> <richard at shockey.us>
> Cc: "Marco bernardi" <marco.bernardi at neustar.biz>; "Andrzej 
> Bartosiewicz"
> <andrzejb at nask.pl>; "Carsten Schiefner" 
> <enumvoipsip.cs at schiefner.de>; <enum-wg at ripe.net>
> Sent: Wednesday, December 15, 2004 9:19 AM
> Subject: RE:: [enum-wg] COCOM & ENUM ...
> 
> 
> Jim writes:
> > I think there's general
> > consensus that end users should not see core telco routing data.
> 
> Here we are at the point: why should they?
> 
> The basic question discussed here is VoIP (aka SIP) on the public
> Internet
> and VoIP (aka NGN - or NWOS - new wine in ols skins - Lawrence) in
> walled gardens.
> 
> If you implement VoIP on the Internet you basically need SIP 
> URIs (AoRs)
> and proxies
> (SVR records in the DNS) to reach this SIP AoRs. If you do not have
> these ENUM in e164.arpa
> Is completely useless. If you have these, you basically do not need
> telcos for routing
> E.164 numbers.
> 
> If you implement VoIP in NGN walled gardens you basically have two
> choices:
> a - they use an extranet (e.g. GRX from GSM-A) to interconnect between
> the walled gardens
> the choice of the routing mechanism (ENUM or something else)
> in this case is purely a decision of the extranet provider 
> (e.g. GSM-A)
> and also limited to the club. End-user will not have access to these
> data.
> What you need here is a mapping E.164 number to operator.
> 
> b - they use the public Internet to interconnect, so they 
> need some kind
> of
> public AoRs (URIs) (e.g. +xxx at be.prov.net )to address the 
> ingress border
> elements
> (these finally have public IP-adresses). So they may use some kind of
> ENUM in a
> commonly agreed upon tree, which could be any root e.g.
> e164.stastny.com)
> Subscribers may may be able to query the tree or not, but it will be
> useless
> because they cannot access the border elements directly.
> 
> The bsaic question here is: do the carriers agree on ONE tree or will
> there be a wood
> of trees? But this is not a big issue if all con-federations have a
> basic rule: any
> participating carrier is obliged to enter only E.164 numbers he is
> serving. Then
> automatically not overlaps exist between the trees (expect carriers
> participating
> in more then one tree) and therefor the trees could be merged easily
> later.
> 
> Back to the original question:
> > I think there's general
> > consensus that end users should not see core telco routing data.
> 
> Normal end-users give a s**t anyway about technical details if making
> a call, so why should the be prevented to see core telco routing
> data. They real problem of telcos is to prevent OTHER TELCOS to see
> their data, primarily to see how many subscribers they really have ;-)
> 
> They funny thing here is that competitors know these facts anyway, but
> it's nice to pretend to have secrets.
> 
> Now we are at the basic problem:
> Is there a way to feed data into a database
> used by competitors to find out which numbers a given telco hosts
> without
> disclosing to the competitors which numbers this telco hosts?
> 
> A cretan says: all cretans are liars. True or false?
> 
> This is the problem we have to solve ;-)
> 
> -richard
> 
> 
> 
>