Re: Big Blue Button

[Jim Garrett]

To those who were able to attend the BBB test meeting last Friday...how
did it go?

I had a conflicting meeting with four parties (including me), which we
conducted with Jitsi using ensemble.scaleway.com.  One person took some
time to figure out how to get her camera to be recognized, but after
some unknown (to me) poking she managed.  Thereafter it was very
smooth.  So that was a success, I think.

Would it be appropriate to create some space on the LibrePlanet wiki to
record summaries of test results?  It could give people guidance based
on actual experience.

-Jim Garrett



On Fri, 24 Apr 2020 18:03:11 -0400
Michael McMahon <address@hidden> wrote:

> Reminder the test is now.
> 
> Best,
> Michael McMahon | Web Developer, Free Software Foundation
> GPG Key: 4337 2794 C8AD D5CA 8FCF  FA6C D037 59DA B600 E3C0
> https://fsf.org | https://gnu.org
> 
> On 4/24/20 12:40 PM, Marc Sunet wrote:
> >
> > Maybe we can meet again next week for those who can't make it today.
> > Might be cool to do this recurrently and maybe bring up some
> > interesting topics :)
> >
> > To summarize, today at 15:00 PST / 18:00 EST:
> >
> > - BBB server: https://testbbb.fsf.org
> >
> > - Jitsi server: https://meet.shellblade.net/LiberationTech
> >
> > As a backup, join #LiberationTech on Freenode, unless someone has a
> > better suggestion.
> >
> > On 4/24/20 9:05 AM, Jim Garrett wrote:  
> >>
> >>
> >> On April 24, 2020 3:19:06 AM EDT, Ruben Rodriguez <address@hidden>
> >> wrote:
> >>
> >>     On 4/23/20 9:17 PM, Jim Garrett wrote:
> >>
> >>         On Thu, 23 Apr 2020 12:12:17 -0400 Ruben Rodriguez
> >>         <address@hidden> wrote:
> >>
> >>             On 4/23/20 11:49 AM, msunet wrote:
> >>
> >>                 Can you meet sometime tomorrow? I am on PST time,
> >> can "leave work" early in the afternoon. 
> >>
> >>             I can join tomorrow too, just not noon to 1pm eastern. 
> >>
> >>         Wow, I'm amazed by all this enthusiasm! I don't have time
> >> to respond as quickly as people are proposing, so if anyone
> >>         would like to go ahead and schedule something, please go
> >>         ahead, broadcast it here, and I'll attend if I can (and
> >>         others will too I'm sure). 
> >>
> >>
> >>     Tomorrow at 6pm EST works for me, and others have suggested
> >> that time. For those who can't make it or want to test the system
> >> at other times, feel free to join https://testbbb.fsf.org
> >>     Note that it is an unauthenticated, single room, demo mode
> >> instance. You can find me anytime on freenode.net IRC as
> >> "quidam" (and at #fsfsys). While we test it I can provide stats of
> >> system load and bandwidth.
> >>
> >>     Please don't use the record feature, as you can't retrieve the
> >>     recordings anyway, and I would have to remove them ;)
> >>
> >>         Let's summarize the results on the wiki page. Which
> >> implies: we should think beforehand of some specific things to test
> >>         out. One specific question I have is: can BBB put all
> >>         participant images on a screen at the same time? Or is
> >> there a maximum?* 
> >>
> >>
> >>     There is no set limit that I know of, but I don't know how the
> >> interface behaves.
> >>
> >>         And when that maximum is approached, does the system show
> >>         signs of stress? It sounds like multiple people have had
> >>         positive experience with BBB, but in truth we haven't yet
> >>         stressed it for the large meeting use case. 
> >>
> >>
> >>     I have only tested it with 3-4 webcams and 20 clients, with
> >> maybe 5 people speaking at the same time. It is supposed to be
> >> good for up to 100-150 people, but in that kind of scenario most
> >> people would have their webcam/audio turned off.
> >>
> >>         Some additional information I think it would be worthwile
> >> to record: - Number of people - Approximate location of server
> >>         and participants (ballpark, not trying to invade privacy,
> >>         just trying to get approximate distance from server) 
> >>
> >>
> >>     The server is in Somerville, MA.
> >>     8 core virtual machine (qemu-kvm) with 8GB ram, running on AMD
> >> Opteron 62xx CPU and KGPE-D16 motherboard with a free software
> >> BIOS (libreboot) and running Trisquel 8 on both the server host
> >> and the VM guest OS.
> >>
> >>     BigBlueButton was deployed with this script I wrote:
> >>     
> >> https://devel.trisquel.info/ruben/communication-install-scripts/blob/master/BigBlueButton.sh
> >>
> >>     It is meant to be used with a frontend (for user management),
> >> we use it with Canvas LMS, but it can be used with several other
> >> platforms.
> >>
> >>         - OS and browser each participant is using - Quality of
> >>         audio: is it scratchy? Does it stutter periodically? How
> >>         often? - Video: how often does it freeze? How is
> >> resolution? Does it align with audio (i.e., do lips move without
> >> relation to audio)? - Someone could purposefully go somewhere with
> >> a weak WiFi connection and see if the system makes reasonable
> >>         trade-offs. (My ideal system would automatically downgrade
> >>         video resolution, then maybe turn off video altogether, and
> >>         at all times preserve audio quality.) - Also, someone could
> >>         purposefully use a laptop's speakers and built-in mic to
> >> test how well algorithms deal with potential feedback - And add
> >>         some background noise while at it. This could be fun. Other
> >>         characteristics? Maybe we could fill out an etherpad
> >>         collectively while we meet. -Jim *Having a maximum number
> >> of videos on screen is an admissible design decision even if
> >>         Zoom does otherwise. The huge corporation at which I work
> >>         uses Microsoft Teams, and it puts at most 4 large images on
> >>         the screen at once. A few additional participants are given
> >>         thumbnails along the bottom, and people beyond that go off
> >>         the screen altogether. When someone speaks, their image
> >> gets promoted to the large screen. This is probably a clever way
> >>         to limit computations and bandwidth for even large
> >> meetings. No one has complained about it yet. 
> >>
> >>
> >>
> >> I have a conflict at that time but please carry on without me! I
> >> look forward to hearing how it went.
> >>
> >> Jim Garrett
> >>
> >> -- 
> >> Sent from my Android device with K-9 Mail. Please excuse my
> >> brevity.