more meeting problems

Mary Fritsch fritsch at mcs.anl.gov
Fri Jul 30 10:24:10 CDT 2004


While mtg in an institution room yesterday, we got dropped from venue 3 
times (lib).  With this abrupt drop, other tasks were running that had to 
be manually stopped each each as well.. causing a big interruption on the 
ANL side.  Everyone else in the meeting (TeraGrid Architecture Meeting) has 
moved back to 1.x version of AG for this meeting series (which is 
troubling).  They were not disconnected from this meeting.

Before my 9:30am tour I wanted to fire up AG in WS to buy some time before 
AM demo, however, the venue client just hung, so I showed some pictures of 
AG mtgs, etc. instead.  At around 10am I was able to connect to vv2/AG Lobby.

I believe all NCSA venues are unicast enabled.  I have used them in the 
past successfully, as well as other bridged rooms, with the simple click to 
unicast method.

Chatrooms within the community have definitely reverted back to 
mud/tkmoo.  I always say hello in the venue textchat, but never get 
responses there.  New characters all the time show up in mud world... 
although folks hang in the Meadow, you still see all the folks in the 
appropriate room in mud space.  This requires me to have a mud character 
and access to the NCSA mud world (venues.ncsa.uiuc.edu/port 7777) for some 
meetings and original Venues world and mud character for 
(ag.mcs.anl.gov/port 7777) for other meetings.  As I have said several 
times, I almost have to setup my chat "plan of attack" before each meeting 
(as it depends on what people can and are using).  Some are IMs, some muds, 
etc.  Sometimes I agree to simply stay connected to Meadow for when text 
chat fails.

Although I am fully aware this is to be fixed in next release, I am afraid 
it is now an untrusted and unused piece of the software.  It also makes 
little sense to keep it there because if you cannot get as far as the venue 
client or you get dropped, you can't communicate with others (which was one 
of the main needs of the back channel of communication).    I have no 
solution, just an observance of behavior from the community to get through 
meetings successfully any way they can.

I do believe the server is more stable, than a few weeks ago, however, we 
pretty much need 100% stability in order for folks to use the venue client 
and 2.x consistently.


At 09:48 AM 7/30/2004 -0500, you wrote:
>things were going fine until it appeared that the multicast to bielefeld 
>was broken.
>
>Tried to swithc to unicast in the ncsa venue, but that didn't work 
>(perhaps firewall there).
>
>Moved to windmer. They switched to unicast, but apparently it didnt' 
>really switch to unicast -- venue client said "unicast" but the tools were 
>using the multicast address. By the time we'd figured this we'd bagged the 
>meeting. again.
>
>I also had the venue lcient in the ws drop me from the venue repeatedly.
>
>we really need a standard text chat mechanism that is completely out of 
>the venue world; we keep getting uttelry cut off. unfortunately this week 
>the guy who I usually talk to over IRC wasn't there.
>


mary fritsch . futures lab . mcs . argonne national lab . 630 252 5297 




More information about the ag-dev mailing list