RE: Session Initiation protocol and videogame consoles

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



> >The idea being running processes
> >(usually games) that don't know their inside SIP.
> >
> >Would it be RTP? Ethernet tunnelling maybe?
> 
> Huh.  Well, if you've got a game that thinks it's just running over
IP, or
> over Ethernet, then you don't really need RTP's timestamps.  On the
other
> hand, you also don't need the reliability of SIP-level messages--and
those
> would be a larger problem than the timestamps, because the game's
protocol
> might include some bits that should be unreliable (e.g., if audio
> streams).

Tunneling protocol data inside SIP is generally considered a bad idea,
as it would dramatically alter the workload of the SIP servers. It is
also a bit weird. If you think about it that would be pretty much like
sending game messages as instant messages. That may work for announcing
your chess moves, but if you try that with a first person shooter
neither the IM provider not the game player are going to like the
result.

The architecturally correct way to use SIP for games is to use a SIP
exchange to initiate a game session, i.e. invite "sip:alice@example.com"
to join a game. The payload of the SIP messages (INVITE and response)
will carry the IP address and port numbers used by the game application.


-- Christian Huitema


[Index of Archives]     [IETF Annoucements]     [IETF]     [IP Storage]     [Yosemite News]     [Linux SCTP]     [Linux Newbies]     [Fedora Users]