Dear Ms. Li Yan,
Thank you for your prompt reply.
Either MCU or a terminal can initiate the call.
In the root-initiated point-to-multipoint case, it's still MCU who should initiated the call since the addditional VC is setup by the calling party, the party who initiated the first VC (Do I understand H.310 procedure right?)
Only initial VC (the call) can be initiated by either MCU or a terminal. An additional VC can be initiated by either MCU or a terminal if the VC is point-to-point. A point-to-multipoint additional VC shall be initiated by MCU.
I am not quite sure that point-to-point additional VC can be initiated by a terminal after H.245 negotiation, though.
We may need to put some guidelines for selecting point-to-point and >point-to-multipoint VC. In the case of one SOT and many ROTs case >shown in Figure 3 (b), MCU shall initiate point-to-multipoint VC >connection as an additional
I agree. But I don't want to make it mandatory.
I supposed that I may have this kind of objection, i.e. not mandatory.
I would like to add something like below in Section 7 or 8.
"MCU shall use point-to-multipoint VC connection for all receive only >terminals. ROT terminals are always receive only.
I understand that it's more efficent to use point-to-multipoint VC connection for ROTs but I am not so sure about making this mandatory. I do not want to exclude the case, where for some reason, MCU does not support Q.2971 but still want to support ROTs. Then in this case, MCU should be able to use point-to-point connection to support ROTS.
I am also afraid that some ROTs may not be able to receive a point-to-multipoint VC properly. In this case MCU should have a fall back mechanism to retry with a point-to-point VC.
Okay. We will change "shall" to "may".
Hideh
Hidenobu Harasaki harasaki@ccm.CL.nec.co.jp Principal Researcher C&C Media Research Labs., NEC Corporation Phone: +81 44 856 8083 Fax: +81 44 856 2232