16 Dec
2004
16 Dec
'04
2:10 p.m.
I spent a few hours looking at the code. It's pretty scary! On the other
hand my c++ isn't as good as my C. It looks like all the input streams
are taken and mixed together and then pushed back to all the sockets
that are connected.
Does one process do this? (I mean there isn't a thread for each person
connected?). In that sense it would be impossible to strip out the packets
that belong to each person as they go back out to those persons. They're
already mixed with everyone elses (in that time slice I suppose).
Besides, analog conference phones don't do that - they just mute the
output as you speak. That wouldn't be a function of sems, that would be
a function of the client...
Just musing :-)
Conor.
Small question on the sems conferencing module. When
you speak you get your
own voice back to you with a small delay. This is rather distracting - is
it possible
to route the voice stream to everyone except the person from which the
stream
came? Perhaps this is not a function of sems but I thought I'd ask.
_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE!
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/