On 1/25/13 10:41 AM, Dmitry Akindinov wrote:
> Hello,
>
> On 25.01.2013 19:23, Paul Kyzivat wrote:
>> On 1/25/13 7:40 AM, Dmitry Akindinov wrote:
>>> Hello,
>>>
>>> On 25.01.2013 16:22, Yaswanth Raparti wrote:
>>>> Hi Brett,
>>>>
>>>> Actually, I gave some random URI to hide the original uri. The original SIP
>>>> URI does not have any '%' . It was encoded though.
>>>
>>> If the Contact URI is syntactically correct and the proxy has inserted
>>> itself as the Record-Route header, then you should not be concerned with
>>> the Contact URI value and meaning.
>>>
>>> If the proxy did not insert itself into the dialog path using
>>> Record-Route headers, then it has a problem.
>>
>> You say "you should not be concerned". That depends on what you want to
>> do. It is true that such a setup will allow *this* dialog to continue.
>> But it won't support anything that needs the contact, like transfer, etc.
>
> True. I stand corrected. I forgot of those transfer scenarios. Sorry for
> possible confusion.
>
> My point was that implementations often try to semantically process a
> SIP URI even when they need a syntactical check at most.

OK, I agree with you on that. If the URI is syntactically correct there 
should be no further complaint about it unless there is a need to use it.

For instance there was another case brought up recently where someone 
was complaining about a Contact of:

<sip:anonymous.invalid;[email protected]:5060;maddr=10.10.10.10;transport=udp;user=phone>

even though it was usable via the maddr parameter.

        Thanks,
        Paul

_______________________________________________
Sip-implementors mailing list
[email protected]
https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors

Reply via email to