Re: AW: AW: [Sipping-tispan] TISPAN requirements, first requiements

["Tom-PT Taylor" <[e-mail]>]

This is a clear statement of REQ-ACR-1 and the surrounding context in

http://people.nokia.net/~miguel/drafts/pre/draft-jesske-sipping-tispan-requirements-02a.txt 


However, what we have actually been arguing over is REQ-ACR-2.

Philip Mart wrote:
> I wonder if I can help you avoid a rathole! If we go back to the original
> source of the requirement we may be able to remove some of this abiguity.
> 
> The base requirement as a stage 1 like text is in DIRECTIVE 2002/58/EC OF
> THE EUROPEAN PARLIAMENT AND OF THE COUNCIL  (Directive on privacy and
> electronic communications)and is available in all community languages from
> the www.europa.eu.int. The relevant text is in Article 8 paragraph 3 and
> states:
> 
> 3. Where presentation of calling line identification is offered
> and where the calling line identification is presented prior to
> the call being established, the service provider must offer the
> called subscriber the possibility, using a simple means, of
> rejecting incoming calls where the presentation of the calling
> line identification has been prevented by the calling user or
> subscriber.
> 
> Calling Line identity is not defined in this directive but the requirement
> to make it available comes from Article 29 of the Universal Service
> Directive (2002/22/EC) which is specifically related to undertakings that
> operate public telephone networks.
> 
> 1) Thus calling line identification can be restricted, if need be, to the
> information conveyed by a TelUri. The requirement is also limited to
> presentation which occurs before the call is established.
> 
> 2) The obligation is on the service provider so the assumption that it can
> be left to a customer's User Agent is not valid.
> 
> 3) We do not need to guess the meaning of anonymous, the requirement is to
> reject calls where the caller has chosen not to provide a calling line
> identity. If this happened because he is connected to the network
> anonymously, and there is also a possibility to provide an identity, then
> the caller has chosen to withold CLI. If there is no network capability to
> provide a CLI the then it is not the caller who prevented the presentation
> of the identity and so the feature should not be activated.
> 
> The ETSI stage 1 requirements in clause 5 of ETSI EN 301 798 states:
> 
> The Anonymous Call Rejection (ACR) supplementary service allows the served
> user to reject incoming calls from users or subscribers who have restricted
> the presentation of their calling line identity according to the CLIR
> supplementary service [3] and [5].
> 
> ACR will reject all calls with CLI marked "presentation restricted"
> according to CLIR. The calls are rejected regardless of the current state
> (e.g. free or busy) of the served user's access. The ACR supplementary
> service shall not reject calls with a CLI marked "presentation restricted
> by network". The served user's ability to originate calls is unaffected by
> the ACR supplementary service.
> 
> The calling user shall be given an appropriate indication that the call has
> been rejected due to the application of the ACR supplementary service.
> 
> NOTE: The method of rejection of anonymous calls is a service provider
> option, and may include the
> functionality where all anonymous calls are forwarded to e.g. a voice mail
> 
> It can be seen that this is a restatement of the directive within the
> context of the ISDN.
> 
> 4) It is here that we get the requirement about the indication. Note that
> the word indication does not refer to a protocol message, instead it is
> just an indication to a person and could be aural, visual or tactile to the
> calling user that a successful call might be made if he behaves
> differently. This sending of information to allow behavioural modification
> is a normal in both the ISDN and SIP.
> 
> At the real "requirements" level I cannot find where any authorised caller
> should be given special treatment. I think that the police are expected to
> release their CLI, after all it could be just a presentation number and
> doesn't even have to be reachable to bypass the ACR mechanism.
> 
> I hope this is one some use
> 
> 
> 
>                                                                                                                                           
>                       "Tom-PT Taylor"                                                                                                     
>                       <[e-mail]>         To:       Miguel Garcia <[e-mail]>                                    
>                       Sent by:                    cc:       [e-mail], "Alexeitsev, D" <[e-mail]>             
>                       sipping-tispan-bounc        Subject:  Re: AW: AW: [Sipping-tispan] TISPAN requirements, first requiements           
>                       [e-mail]                                                                                                         
>                                                                                                                                           
>                                                                                                                                           
>                       25/08/2005 14:59                                                                                                    
>                                                                                                                                           
>                                                                                                                                           
> 
> 
> 
> 
> You are missing a case.  The distinction is whether the caller is
> anonymous to the network as well as to the called party.  So one of the
> ACR cases is when the network has the identity of the caller but will
> not disclose it due to operation of the Privacy service.
> 
> Miguel Garcia wrote:
> 
>>So what is an anonymous user? An anonymous user is one for which his
>>identity is not determined. The reason why a user is anonymous is
>>another story (the user want, or the network can't provide an identity),
>>but they are anonymous in both cases.
>>
>>/Miguel
>>
>>Tom-PT Taylor wrote:
>>
>>
>>>I think your requirement is not quite accurate even at this level of
>>>abstraction.  Roland has identified two different cases.  One is your
>>>"certain anonymous users", but the other is the case where the user
>>>identity is unable due to the network's inability to provide it.
>>>
>>>I guess I can see your point about pushing too hastily to mechanism.
>>>Roland actually identified the information that might be used to
>>>resolve the issue at the rejection point: on the one hand, the
>>>identity of the caller if present in a P-A-Id, and on the other hand,
>>>the absence of a P-A-Id.  It may be that this is all that is needed,
>>>and no further discussion is required in SIPPING.  (Of course, that's
>>>assuming that the number of authorized caller identities any one
>>>rejection point has to keep track of is limited and provisionable.)
>>>
>>>Miguel Garcia wrote:
>>>
>>>
>>>>Tom:
>>>>
>>>>I agree with you in the essence, but you are referring to a solution,
>>>>not to a requirement. Essentially, whenever you need to speak about
>>>>an "indicator", it raises an alarm in my head that you are speaking
>>>>about a possible solution rather than a requirement.
>>>>
>>>>So the requirement is tha there are certain anonymous users that are
>>>>not subject to be filtered by ACR. Then we can further discuss the
>>>>solution, which is probably going in the direction you suggest.
>>>>
>>>>/Miguel
>>>>
> 
> ...
> 
> 
> _______________________________________________
> Sipping-tispan mailing list
> [e-mail]
> https://www1.ietf.org/mailman/listinfo/sipping-tispan
> 
> 
> 
> 
> 
> 
> 


_______________________________________________
Sipping-tispan mailing list
[e-mail]
https://www1.ietf.org/mailman/listinfo/sipping-tispan