[ejabberd] empathy jingle and ejabberd

Peter Saint-Andre stpeter at stpeter.im
Tue Sep 16 04:11:24 MSD 2008


Michael Flaig wrote:
> Hi,
> 
> as I understand jingle there are 2 possible ways to transport video and
> voice. The first is similar to sip, xmpp is used to do signaling and
> chat but audio and video goes directly from client to client using RTP
> Protocol. Another, newer mode is to have RTP inline in jingle, which
> routes video and voice through the jabber server, of course puts more
> stress on the server and increases latency, but allows to use jingle if
> there is no direct connection to the internet.
> Anyway, the first mode is the common one and will be finalized soon (as
> in XEP going to draft state). If you have 2 devices behind NAT you need
> a STUN server, as it is the case with SIP.
> 
> Hope I did not misunderstand some of the jingle stuff and wrote
> incorrect things here... if so, I'm sorry and would appreciate
> correction :-)

We really don't encourage people to do voice or video using any of the 
in-band options. That might be appropriate for file transfer (you want 
to make sure the file gets there, but time is not critical), however the 
kind of performance you'll get from in-band bytestreams will not be 
acceptable for voice and certainly not for video. Plus your jabber 
server might fall over from the load. :)

> So shot: jingle does not required anything special from the server.

True.

> If you want to get to the bottom of you problem I suggest looking into
> the packets empathy send out with a packet tracer like wireshark.

Good idea.

Peter
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 6751 bytes
Desc: S/MIME Cryptographic Signature
URL: <http://lists.jabber.ru/pipermail/ejabberd/attachments/20080915/304f59c3/attachment.bin>


More information about the ejabberd mailing list