Sidharth,

You have found an error in the basic call flows document that I will 
fix.  In Message F10 of flow 3.7, A does change the port, so the version 
number in the SDP should be incremented.  I will make this change to the 
document.  Any time the SDP has changed, the version number should be 
changed as well.

Also, for questions about SDP and call hold, take a look at the Service 
Examples draft 
(http://www.ietf.org/internet-drafts/draft-ietf-sipping-service-examples-02.txt) 
since it shows (hopefully) the correct behavior in a few flows.

Thanks,
Alan Johnston
sip:[EMAIL PROTECTED]

At 02:20 PM 9/4/2002 +0530, [EMAIL PROTECTED] wrote:



>Hi Senthil,
>
>Thanks for your response. Please find my comments inline
>with "SID>>" preamble.
>
>Cheers,
>Siddharth.
>
>-----------------
>Siddharth Toshniwal @ Hughes Software Systems
>http://www.hssworld.com
>
>
>1. In case the UAC sends SDP in INVITE with the session-id
>    in the "o=" line as (say) 2845844566, is it a MUST for
>    the UAS to return the same session-id in the answer?
>    The offer/answer-02 draft allows the UAS to have an
>    independant version number, but is silent about the
>    session-id field. Can the UAS use a different session-id
>    and use the same session-id henceforth? The call-flows
>    draft (draft-ietf-sipping-basic-call-flows-00.txt) does
>    seem to indicate this.
>
>    [ The example on section 10 is an clear indication
>      that the session ids can be changed. Reason the o=
>      line needs to be changed.]
>
>SID>> The example shows this, yes. But the call-flows do
>SID>> have an occasional omission/error. So I wanted to
>SID>> confirm :) Read on please, and the intent of this
>SID>> question will be clearer.
>
>
>2. Same offer, but a different answer
>    ----------------------------------
>    Should the 'version' field of the "o=" line in the answer
>    be incremented even if the offer was unchanged and the
>    answer is different from the original answer in
>    (a) only the ports?
>    (b) only a change in IP address?
>    (c) change (addition/subtraction) in accepted codecs?
>
>    [ If I change my SDP I would like to version it to keep
>      track of recent offer/answer I have made . The answer
>      is yes, if the answerer decides to give an different
>      SDP he would have an different version.]
>
>SID>> Again, if the call-flows are correct, then the version
>SID>> need not be incremented for a change in the port number.
>SID>> The example that begins on Page 62 of the latest call-
>SID>> flows draft (draft-ietf-sipping-basic-call-flows-00.txt)
>SID>> seems to indicate that. Please refer to the SDP message
>SID>> in the steps F1 and F10 of the call-flow (Section 3.7)
>SID>> I mentioned. The user A has changed port of the stream,
>SID>> but not updated the version in the SDP. Question arises,
>SID>> is this a nit in the call-flow draft or is this the way
>SID>> its supposed to be? The rest of my questions above were
>SID>> extensions to the doubt.

>3. Different offer, but same answer
>    --------------------------------
>    If the offer is different from the previous offer by
>    the UAC (he's trying to add some codecs, say); but the
>    answer that the UAS generates is the same as the previous
>    answer (all additions that were attempted are rejected),
>    then should the UAS increment its version in the o= line?
>    If it does, is it an error?
>
>    [ No you have to increment version from the previous version
>      which has been answered for the previous offer ]
>
>SID>> Why so? In case the caller wants to put our client on hold,
>SID>> and we do not put the caller on hold in turn, then the
>SID>> answer involves no change to the session data. In such a
>SID>> case is it necessary for the server to increment version?
>
>
>I feel it would be good if the offer/answer draft was
>made more explicit regarding the handling of the o= line.
>
>   [ I personally feel RFC 2327 is clear on version usage]
>
>   [<version> is a version number for this announcement.  It is needed
>    for proxy announcements to detect which of several announcements for
>    the same session is the most recent.  Again its usage is up to the
>    creating tool, so long as <version> is increased when a modification
>    is made to the session data.  Again, it is recommended (but not
>    mandatory) that an NTP timestamp is used.]
>
>SID>> What constitutes a change to the "session data"
>SID>> is my point. Is it a change in codecs, or a change
>SID>> in ports or a change in IP address where the stream
>SID>> has to flow? Ideally, it should be any combination of
>SID>> the 3, I believe.
>
>
>
>_______________________________________________
>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