Hi Subhash,
The criteria itself is open to implementations but I assume based on the
authentication information proxy need to somehow identify a list of
identities that could be validly associated with that user. For ex, this
could either be a database of its own or a database provided by other
component on which it relies. Otherwise malicious UACs might indicated
false identities in PPI. 
In the example if PPI is something like P-Preferred-Identity: "John Doe"
<[EMAIL PROTECTED]>, proxy need to verify that this is not a valid
identity for user fluffy and either it rejects this requests or inserts
[EMAIL PROTECTED] or [EMAIL PROTECTED] The reason the example includes
[EMAIL PROTECTED] might be that the domain being forwarded to is
vovida.org and there might be an implicit trust relationship between
those two.

There are some valid reasons where proxy might need to ignore PPI and
insert a totally different PAI which has no semblance to PPI at all.
When forwarding requests across domains, proxy might need to insert a
PAI which a user is not aware of or might need to provide a different
PAI to the forwarding domain to provide certain features.


-Venkat 

-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] 
Sent: Friday, May 14, 2004 10:21 AM
To: Goud, Venkat Ramana (Venkat) ** CTR **
Cc: [EMAIL PROTECTED]; [EMAIL PROTECTED];
[EMAIL PROTECTED]; [EMAIL PROTECTED]
Subject: RE: [Sip-implementors] Interpretation of 'hint' in
thePreferred-Identity






Hi Venkat,
      My question is how does the proxy recognize if
the PPI is a valid identity for the user ? From the
example, it seems like
P-Preferred-Identity: "Cullen Jennings" <sip:[EMAIL PROTECTED]> results
in the proxy inserting:
P-Asserted-Identity: "Cullen Jennings" <sip:[EMAIL PROTECTED]>

In this case, how did the proxy decide that the PPI was a valid or
invalid identity ? Ideally it should have inserted [EMAIL PROTECTED]
but it seems to have inserted [EMAIL PROTECTED] Is it correct to assume
that it did a _full_URL_comparision_; did not find a match; and hence
inserted the closest identity it could find to the one suggested in the
PPI ? Again, in deciding "closest identity" what was the criteria - the
PPI header does not allow the client to suggest which part of the header
is to be used as the hint (display-name or userinfo part or
whatever)...is that left open to implementations ?

Thanks,
Subhash.



 

             "Goud, Venkat

             Ramana (Venkat)

             ** CTR **"
To 
             <[EMAIL PROTECTED]>          <[EMAIL PROTECTED]>,

 
<[EMAIL PROTECTED]>  
             05/13/04 06:28 PM
cc 
                                       <[EMAIL PROTECTED]>,

                                       <[EMAIL PROTECTED]>,

                                       <[EMAIL PROTECTED]>

 
Subject 
                                       RE: [Sip-implementors]

                                       Interpretation of 'hint' in

                                       thePreferred-Identity

 

 

 

 

 

 





I think section 6 is pretty clear on this. PPI is only a
suggestion/request so I guess its up to the proxy as to whether to use
the hint or not.

If there is a P-Prefferred-Identity header:

1) If the proxy recognizes it as a valid identity for that user and if
there is no local policy for the proxy suggesting some other PAI, then
it can use the one in the PPI header.

2) If the proxy does not accept the PPI as valid identity for the user
then it may either reject that request or use a PAI constructed based on
its local policies.

-Venkat




-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] Behalf Of
[EMAIL PROTECTED]
Sent: Thursday, May 13, 2004 8:33 AM
To: [EMAIL PROTECTED]
Cc: [EMAIL PROTECTED]; [EMAIL PROTECTED];
[EMAIL PROTECTED]
Subject: [Sip-implementors] Interpretation of 'hint' in
thePreferred-Identity






Hi,
      RFC 3325 states that a proxy may use the P-Preferred-Identity in
an incoming request as a hint to choose among multiple identities that
are configured for a user. The example section shows an example where
the user suggets that he prefers to use the identity [EMAIL PROTECTED]
But the proxy inserts [EMAIL PROTECTED] in the forwarded request.
      The RFC does not clearly specify how the proxy should interpret
the 'hint' that is provided. From the syntax of the P-Preferred-Identity
and the P-Asserted-Identity it seems that one should do a URL comparison
to choose the correct PAI header. But from the example, it seems like
either the display-name or the userinfo part is being used to choose the
correct PAI header.
      Can someone provide some clarity or point to discussions which
indicate how the 'hint' should be interpreted. Is it that no rules have
been laid down so that it is left to individual implementations ? In
that case, could implementors/authors suggest as to how one would
normally interpret the hint i.e. by comparing which part of the
P-Preferred-Identity header.

Thanks in advance,
Subhash Nayak
Hughes Software Systems
http://www.hssworld.com

_______________________________________________
Sip-implementors mailing list
[EMAIL PROTECTED]
http://lists.cs.columbia.edu/mailman/listinfo/sip-implementors



_______________________________________________
Sip-implementors mailing list
[EMAIL PROTECTED]
http://lists.cs.columbia.edu/mailman/listinfo/sip-implementors

Reply via email to