We currently have an open Stan meeting each Thursday 11am through Google hangout.
One concern that have arisen is that meetings can go on awhile on topics that aren’t so interesting to most of the people in the (virtual) room.
We’ve discussed two solutions to this problem. First, at the start of the meeting we do a very quick check-in where each participant has no more than 2 minutes to present news. Second, we become more formal in setting up the agenda for each week’s meeting, so that people know ahead of time what is the primary topic of discussion.
It might be helpful to consider three kinds of discussions:
-
Decisions. Something comes up, a decision large or small, and we discuss in the meeting. We’re not talking about a formal voting process but rather open-ended discussion with a decision goal.
-
Stan discussions. Items related to language, implementation, etc., not specific decisions but important topics that would benefit from open discussions.
-
Research discussions about Bayesian statistics, machine learning, algorithms, applications, etc.
I don’t have strong opinions on how to organize Stan meetings, I just thought the above classification could be a helpful way to think about things.