[AG-TECH] Conference XP vs AG

[Patrick Bristow]

I'll jump in here, as I think Chris is busy (and doing things that
involve my yearly review, which I'd prefer he keep working on! ;)

So I also don't know a lot about the ability to tie services together
across AG machines, but I think I understand the basic functionality.
In my understanding, the only difference between what AG allows and
simply putting two computers in the venue is that with the AG, you're
more clearly representing that the machines belong to only one logical
"participant" or room.

CXP will happily allow you to just setup two computers and put them each
in the venue, both with a set of cameras attached, etc.  However, there
are no provisions for tying the two logically together, and we won't
make the experience any easier or harder for you.

For Derek, you can easily override the identity of the participant by
one of two methods:

* From the commandline/shortcut - e.g. "cxpclient.exe -e client at ms.com"
* By adding this entry in the CXPClient.exe.config file, 
<add key="MSR.LST.ConferenceXP.Identity.Identifieroverride"
value="another email address"/>

I think that by just running the client separately on both machines &
doing the Identifier Override as necessary will give a very similar
experience to the way the AG makes the two machines "appear as one".
Perhaps the only odd thing is that you'll see both/all machines in the
venue, but you can make more clear by just setting the profile name
field to be similar, such as "Argonne, Room 112, #1" and "Argonne, Room
112, #2".

Please correct me if I'm wrong about the AG functionality.  I'm not an
experienced AG user.

Thanks,
Patrick

-----Original Message-----
From: owner-ag-tech at mcs.anl.gov [mailto:owner-ag-tech at mcs.anl.gov] On
Behalf Of Todd Zimmerman
Sent: Wednesday, August 31, 2005 10:25 AM
To: Chris Moffatt
Cc: ag-tech at mcs.anl.gov
Subject: Re: [AG-TECH] Conference XP vs AG

Hi Chris,

In our case, this capability is very important for a couple of reasons.
First is the basic reason
that has been described already - spreading the load.  Our theatre node
drives 4 displays and has
three captures.  While I may be able to pull that off with one machine,
why would I do that when I
can easily split the tasks to two machines?  This frees up the display
machine to possibly do other
tasks also (running remote visualization software etc) without getting
bogged down.

Secondly, I think you are correct that currently the main purpose is the
ability to 'tie a/v devices
across multiple machines'; however, the nice part about the
servicemanager/services infrastructure
is that it is completely extensible.  It doesn't have to be about
typical a/v at all.  Instead of
having an audioservice or a videoservice, imagine having a 3D
visualization service - which may
require the control of a specialized display or a specialized set of
devices etc.  With this
infrastructure, we are capable of building our own specific services
that are controlled easily and
centrally.

Just my $0.02... (Actually, that's Canadian dollars too, so its probably
only worth about a $0.015
to you... ;-) )


Todd

-- 
Collaboration & Visualization Technician
WestGrid - www.westgrid.ca
IRMACS - www.irmacs.com
Ph. 604.268.6979
Todd Zimmerman - toddz at sfu.ca