Re: Few queries in regards to MSRP
Ben Campbell <[email protected]>
| Newsgroups | gmane.ietf.simple |
|---|---|
| Message-ID | <[email protected]> |
Hi, see inline: On Jan 25, 2012, at 10:33 AM, prasun bheri wrote: > Hello Group, > > I have these following queries with respect to MSRP: > > Section 5.4 says "active endpoint MUST immediately issue a SEND request" however no timeout period is specified so hoping its application specific, I would like to know on what factors does this timeout value depend on? I am using MSRP to deliver and/or receive MMS messages, so any suggestions on initial timeout value in this case? What would be the right thing to do when initial message is not received immediately? There's not a timeout per se for this. The idea is that, if the active party doesn't have an actual message ready to send when it sets up the connection, it should send a dummy message. The passive party should not assume anything about the identity of the active party until it receives at least the header fields for a Send request. Afterwards, it can determine the connected party by comparing the received MSRP URI to the one it handed out in the SDP. If it doesn't get something in a timely matter, it could drop the connection (and tear down the SIP dialog)--but timely manner here probably means minutes, not seconds. > In case of active end point, if no payload is immediately available, protocol requires that active endpoint sends an empty send message with byte-range 1-0/0, Could this send message have same message-id as that of payload's message-id that follows this initial send? My understanding is that both of these should have different message id, but I would like to hear your comment on this. Since it would not be part of any subsequent message payload, it should have a different message-id. > Once a session is created, is there a way to know if this communication is only one way or half duplex or full duplex? This would normally be part of the session setup--i.e. in the SDP exchange. See section 8.9 of RFC 4975. Hope this helps! Ben. _______________________________________________ Simple mailing list [email protected] https://www.ietf.org/mailman/listinfo/simple