Hi,

Regarding the value for AttachReqAns.role:

    /*
    * RELOAD 5.5.1.8. Role Determination
    *
    * "The roles of controlling and controlled as described in Section 5.2
    *  of ICE are still utilized with RELOAD.  However, the offerer (the
    *  entity sending the Attach request) will always be controlling, and
    *  the answerer (the entity sending the Attach response) will always be
    *  controlled."
    *
    * RELOAD 5.5.1.13. Sending Media
    *
    * "Consequently, once ICE processing completes, the agent will begin
    *  TLS or DTLS procedures to establish a secure connection. The node
    *  which sent the Attach request MUST be the TLS server.  The other
    *  node MUST be the TLS client."
    *
    * Therefore, according to RFC-4145, the offerer (ICE controlling)
    * MUST take the 'passive' role and the answerer (ICE controlled)
    * MUST take the 'active' role.
    */

I want the group to confirm this, because it seems counter-intuitive. If this
is true, then I suggest the following change to the definition of 'role' in
section 5.5.1.1 to help implementers avoid a critical mistake:

  role
      An active/passive/actpass attribute from RFC 4145 [RFC4145].
      This value MUST be 'passive' for the offerer (the peer sending
      the Attach request) and 'active' for the answerer (the peer
      sending the Attach response).

Thanks

--Michael
_______________________________________________
P2PSIP mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/p2psip

Reply via email to