[Spce-user] actual device that initiated BYE?

[Stefan Sayer]

o Jon Bonilla (Manwe) on 08/15/2012 01:16 PM:
> Neeed to retest some day and check what oej says in the ticket, but I'm pretty
IMHO Olle is mixing up things in the comment of the new ticket. 
Especially, the part about the Require indicating applying SST is 
wrong - it's of course the Session-Expires header that indicates how 
SST are applied, and if the UAS says 'refresher=uas' then the UAS is 
supposed to do the refresh.

The bit regarding the Require and why it's a MUST is even explicitely 
explained:
"""
If the refresher parameter in the Session-Expires header field in the
    2xx response has a value of 'uac', the UAS MUST place a Require
    header field into the response with the value 'timer'.  This is
    because the uac is performing refreshes and the response has to be
    processed for the UAC to know this.
"""

If asterisk doesn't compute the refresher role properly, it should not 
add Session-Expires header in the first place (it can still do 
refreshes as it likes, regardless of what intervals are negotiated or 
not).

Note also that in newer (1.5+) SEMS, the Require/Supported headers are 
set properly.

What could be useful in such situations where we have to deal with 
such other implementations, though, could be a third mode where 
refreshes are just done always by the B2B, and no SST is negotiated.

Stefan