Re: MPCS, NM 2.1, and CU 3.1

Jason Williams (
Tue, 21 Jul 1998 09:20:32 -0500 (CDT)

On Tue, 21 Jul 1998, Gary Dietz wrote:
> 1) NetMeeting Participant 1 in --> MPCS conference (configured as H.323)
> 2) NetMeeting Participant 2 in --> MPCS conference (configured as H.323)
> 3) NetMeeting Participant 3 in --> MPCS conference (configured as H.323)
> 4) CU 3.1 Part. 1 in same conference (sending H.263/G.723)
> 5) CU 3.1 Part. 2 in same conference (sending H.263/G.723)

Know anything about what Scott was telling me about the 2.X clients using
the H.263 codec to interoperate with Netmeeting clients on an MPCS server?

I still need to test some things out concerning Cornell. I'm not sure if
MPCS would convert the Netmeeting users to be able to be seen by non-RTP
CU clients (such as the Cornell client which can see H.263 just fine as
long as the codec is installed). The same with the Mac Cornell client
that's in testing stages. I believe the last time I really tried this,
the non-3.X clients couldn't even see the Netmeeting user as a participant
on the reflector at all. Probably Netmeeting and MPCS using RTP (dunno
though...does H.323 use RTP?)

> Chat will not interoperate between CU and NM (yet).

I hope this can be fixed at some point in the future..perhaps when chat
becomes standardized. CU handles chat a lot more cleanly than Netmeeting
(in my opinion) as well as Internet Phone. iVisit's solution is nice.
Perhaps in a year or two, iVisit will become H.323 compliant and we can
all be one happy family. :)

> The point of MPCS and NetMeeting *today* (or MPCS and ProShare for that
> matter) is really for NetMeeting-only sites to be able to distribute audio
> multipoint and have the conference control capability that MPCS provides.

Is there any word of a mulitpoint native version of NetMeeting? I heard
about this being tossed around about a year ago on comp.dcom.videoconf. I
believe the H.323 solution to multipoint connections is with multicast, so
it still leaves out modem users.

> In the near future, some really cool tools to allow client based selection
> of received video in 3rd party H.323 clients. Neat stuff.

Oh? Do tell :)

> Also, as a really cool point, when we have an Intel Business Video
> (Proshare) client sending video through MPCS on a LAN, using either
> CU-SeeMe or NM as endpoints, we can *RECEIVE* the hardware assisted video
> of the ProShare client at over 18 frames per second (and approaching 30 fps
> on the right hardware).

Given a LAN, you can do just about anything :)
Unfortunately, the people I know on CU aren't on LANs. :)

--    * Jason Williams -- Austin, Tx.  |     |       * University of Texas at Austin  | ___ |         * BS Computer Science             \_|_/
*************** **************|