FW: [Fwd: Windows software for connecting to an AG conference]

[Ivan R. Judson]

..........
Ivan R. Judson .~. http://www.mcs.anl.gov/~judson
Futures Laboratory .~.  630 252 0920
Argonne National Laboratory .~. 630 252 6424 Fax
 

> -----Original Message-----
> From: Philippe Galvez [mailto:galvez at hep.caltech.edu] 
> Sent: Monday, June 10, 2002 7:57 PM
> To: judson at mcs.anl.gov
> Cc: Rowe at bmrc.berkeley.edu; openmash-developers at openmash.org; 
> warner at cats.UCSC.EDU; mantey at soe.ucsc.edu; 'Philippe Galvez'
> Subject: Re: [Fwd: Windows software for connecting to an AG 
> conference]
> 
> 
> 
> Sorry to be a bit late but was 100% busy by others stuff. 
> There are several notions which have been mentioned there. I 
> will try to catch few and provide some inputs
> 
>     o OpenMCU and AG venue: I am not sure what is the 
> intention here. Is it to connect H.323 to AG ? It is not as 
> obvious as it looks like. Different audio compression (Linear 
> 16bits/16Khz versus G.711) or multi video versus one video at 
> a time (H.323). OpenMCU is surely not a good schema for 
> distributed model.
> 
>     o VRVS/AG gateway: There are no clients (or VRVS 
> clients), everybody used what he wants (Mbone, or any H.323 
> devices). Right, the server is not open source (Caltech 
> policy, nothing I can do about it). There are no hand 
> configurations, an advanced web administrative interface is 
> used to interact with any VRVS network servers when necessary.
> 
>     o We should make the difference between how the VRVS 
> architecture is design and how it is implemented. Today, all 
> the VRVS network server (42 world wide) can perform an 
> AG/H.323 bridge. But we allow only the ANL bridge to do it 
> because it is a dedicate machine and to not want to perturb 
> all our current production meeting (average 15 per days) with 
> additional high multicast traffic grabbed by the machine.
> 
>     o It take 2 seconds to add new multicast address or new 
> venue. The question is not if it is possible is more how/what 
> do we present as UI to the user to do it. We could have 10 
> different answer is we ask 10 different users.
> 
>     o The current system  is purely design to be distributed. 
> It is simply software. The new version currently in testing 
> phase (and very soon in
> production) is now able to transcode Linear16/16Khz to PCM. 
> So, we have a fully interactivity between H.323 and AG. And 
> we will work on some other interesting features that will be 
> add to this.
> 
> Well, I hope it is clear at least on what VRVS/AG is exactly.
> 
>         Regards,
>         Philippe
> 
> "Ivan R. Judson" wrote:
> 
> > Hey,
> >
> > Only limited anecdotal experience.  My impressions (I'm 
> sure Phillipe 
> > can correct me if I'm wrong) are only these two:
> >
> > 1) the clients are free, the server is not available (which runs 
> > counter to our community model, which is moving even more 
> distributed 
> > as we get into peer-to-peer stuff)
> > 2) it's a lot of processes that look like it takes a lot of hand 
> > configuring to get it going, it might run smoothly without any 
> > problems for long periods, I don't know.
> >
> > Like I said I haven't used it much, it is running on some 
> ANL machines 
> > to bridge people into the venues now (for some venues, probably not 
> > the institutional ones).
> >
> > When venues servers become distributed (allocating media 
> channels out 
> > of locally scoped multicast addresses dynamically) many of these 
> > systems for bridging will have to be significantly altered 
> unless they 
> > are designed to handle that now. We're not that far from 
> being able to 
> > deploy the new venues services and it'd be a shame to have 
> "connected 
> > communities" have another hurdle to integration.
> >
> > --Ivan
> >
> > -----Original Message-----
> > From: owner-openmash-developers at bmrc.berkeley.edu
> > [mailto:owner-openmash-developers at bmrc.berkeley.edu] On Behalf Of 
> > Lawrence A. Rowe
> > Sent: Monday, June 03, 2002 7:44 PM
> > To: openmash-developers at openmash.org
> > Cc: warner at cats.UCSC.EDU; mantey at soe.ucsc.edu; Philippe Galvez
> > Subject: [Fwd: Windows software for connecting to an AG conference]
> >
> > Hi Ivan -
> >
> > "...integrate OpenH323's OpenMCU with the AG Venues..."  
> sounds like 
> > an interesting project.  i'll add it to my list of "projects for 
> > students" and we'll see if anyone is interested.
> >
> > we're looking at VRVS as a bridging technology to H.32x and other 
> > systems. do you have any experience with that?
> >         Larry
> > --
> > Professor Lawrence A. Rowe          Internet:  
> Rowe at BMRC.Berkeley.EDU
> > Computer Science Division - EECS       Phone: 510-642-5117
> > University of California, Berkeley       Fax: 510-642-5615
> > Berkeley, CA 94720-1776            URL: 
> http://bmrc.berkeley.edu/~larry
> 
> 
>