Re: CU Video + Audio

Jason Williams (streak@ccwf.cc.utexas.edu)
Wed, 24 Sep 1997 02:38:44 -0500 (CDT)


On Wed, 24 Sep 1997, Rolf Hemmerling wrote:
> So we must blame it on the refs.. and not on the bandwidth needed by
> thee clients... I thought so, but I couldn=B4t prove. With direct
> connections, I had less problems and successful audio contacts (if I
> stop video just for audio sending).

This isn't a fair comparison really. With direct connections, there's no
potential for multipoint conferencing. The more clients that are on the
reflector, the less likely it is audio will work well. With only one
person there though (ie: a DC) audio is likely to work much better simply
because you don't have more video windows open.

> The real problem with the refs: 90% of connections are with 10% of the
> refs, the rest of the refs is empty (so that they are free for personal
> dates, but it is not useful to wait for random contacts like it is usual
> for CU-SEEME folks).=20

True...that's the way things have always worked though. I've often
thought about having a better way to coordinate things on a mass scale.
Certain reflectors are always full, but then most reflectors are usually
always empty. It makes me wish there was a way to distribute load on a
reflector so that it's more evenly distributed. If you have a group of
reflectors connected together, connecting to one address would actually
connect to one with the least amount of load (fewer participants). This
evens the load out a bit, but also prevents you from seeing who you might
want to see.

The idea for me on this comes from the login to a local unix system. It's
a cluster of unix machines and when you enter a generic machine name, it
performs load balancing to determine which machine you get to login to.
It helps even the loads out as long as people use the generic machine name
to get to it. If lots of people go to the machine specifically, it
doesn't do much good.

> Are there many refs just connected with 64 KBit (ISDN) or so with the
> net ? If this is true, I don=B4t wonder about bad video and audio
> connections.

I don't know of ANY that are connected via 64kbps ISDN...I used to know of
one reflector connected up thru 128kbps ISDN..but it didn't last long.

All the reflectors I've seen have at least a T1 connection to the
internet.

--
streak@ccwf.cc.utexas.edu    * Jason Williams -- Austin, Tx.  |     |
streak@mail.utexas.edu       * University of Texas at Austin  | ___ |
streak@cs.utexas.edu         * BS Computer Science             \_|_/
*************** http://ccwf.cc.utexas.edu/~streak/ **************|