Re: [BEHAVE] STUN/TURN message type bits

Jonathan Rosenberg wrote:
> I think its wrong. The reason is that the class is only two of the 12 
> bits. So I think its:
> 
> #define IS_REQUEST(msg_type)       (((msg_type) & 0x0110) == 0x0000)
> #define IS_INDICATION(msg_type)    (((msg_type) & 0x0110) == 0x0010)
> #define IS_SUCCESS_RESP(msg_type)  (((msg_type) & 0x0110) == 0x0100)
> #define IS_ERR_RESP(msg_type)      (((msg_type) & 0x0110) == 0x0110)

thanks for answering my questions, Jonathan, and you're right with 
the bit-fields. Using 0x0110 mask rather than 0x0FF0 gives us 
possibility to have more than 16 STUN methods.

>>    0x0004  :  Send Indication
>>    0x0115  :  Data Indication
>>    0x0118  :  Connect Status Indication
> 
> Yes, the TURN values are wrong. I discussed this at the IETF meeting, 
> that the proposed bitfield approach worked well for existing STUN 
> attributes but would be wrong for the new indications that were recently 
> added to TURN. Those need to be changed in the revision of TURN.

So until the new revision is released, I think we can assume that 
these are the values (?):

     0x0014  :  Send Indication
     0x0015  :  Data Indication
     0x0018  :  Connect Status Indication

thanks
  -benny

-- 
Benny Prijono
http://www.pjsip.org

_______________________________________________
Behave mailing list
Behave@ietf.org
https://www1.ietf.org/mailman/listinfo/behave