[all][Virtual PTG] Notes from the 'Running Virtual Meetups' session

Artom Lifshitz alifshit at redhat.com
Tue Apr 7 13:54:11 UTC 2020

Hey all,

Last week, Red Hat hosted a 'Running Virtual Meetups' virtual session,
with presenters from both Red Hat and others in the industry. I
attended, hoping to gleam potentially interesting insights into how we
can run our virtual PTG. Here are my notes:

In real PTG, the physical and geographical aspects of it make it much
easier to separate from our usual work. The usual calls, tickets,
emails, whatever, take a back seat due to the constraints of a PTG
setting. In a virtual PTG, with us sitting at our usual desks, this
will become much harder.

This leads into the next point - participant engagement. Having
technical conversations for 8 hours a day is already hard enough when
you're in the same room as the people you're conversing with. When
those people are pixels on your screen, it becomes even harder. The
session I attended was geared more towards presentations than
PTG-style roundtable discussion, so I'm not sure that the proposed
mitigations would apply to us. For instance, live polls make no sense
in a PTG context.

One interesting suggestion was to do physical movement between
topics/sessions. The online attention span of a human is estimated to
be around 10-15 minutes, so having participants stand up and walk
around their desk/kitchen/hallway 4 times an hour could be something
worth trying.

Another way to keep folks engaged is to actually do things besides
talking. In our context, this could mean hacking on some POC code
together, or collaboratively writing a spec. Again, the session I
attended was geared more towards presentations, so the suggestion was
to do live debugging. It's actually more engaging when things break
and you fix them as a group, than when a presentation goes off without
a hitch.

That was about it. There was also a list of potentially useful tools,
that I'll present here as a link dump because I've never used any of


The danger is that too much tech brings complexity that's difficult to
manage, so sticking to simple well-known tools can also be an

Hope this is useful, cheers!

More information about the openstack-discuss mailing list