Why do meeting room times overlap?

It seems strange to me that we have a limited number of meeting rooms in Matrix, especially when most “meetings” are just text chat. This unnecessarily causes:

  • Teams to sometimes step on each other when they shuffle their meeting times.
  • Confusion about which room the meeting is in.
  • An inability to scroll back and see e.g. the last several meetings (or even just the most-recent meeting) from a particular group.

Can someone explain how we got here / why we’re still here?

It seems to me that we could instead do either:

  • One meeting room per group, e.g. #admin-meetings, #docs-meetings, etc. (This would be my preference.)
  • Groups just hold their meetings directly in their team channel, e.g. in #admin or #docs.

If the reason is simply, “It’s a pain to create and configure everything” then maybe it’s something we could pick at like with the Forge migration? I’d be happy to help (or even better, automate!), though presumably I would need some privileges.

Having a few defined meeting rooms allows for people to lurk in the room and be pinged if they’re needed for a meeting that they don’t normally attend, or to just see what’s going on. The meeting room concept broadens transparency across teams in Fedora. Plus it allows non-meeting conversations to continue in the team’s main channel. This was established when IRC was our real-time chat and it’s worth considering if that model still holds now that we’re primarily on Matrix.

There is also some overhead on the admin side. It’s less about creating the channel and more about making sure the bot is working in the channel.

3 Likes

There was also a longer discussion in Should we have meetings in dedicated meeting rooms, or in the team's own channel? about why we have separate meeting rooms. The Go SIG, for example, used to do the meetings in our own room and now use the meeting rooms. I think there are already three or four meeting rooms, but if that’s not enough to resolve conflicts, it should be relatively straightforward to create another one.

2 Likes

Thanks for the feedback, Ben.

Both of these are still true in the “room per group” model.

Bonus:

  • You get to pick which conversations you’re lurking in, instead of “all of them” / “a random subset”.
  • You can configure your Matrix client to alert you when e.g. an Infra meeting is happening.

This is still true with the “room per group” model. If someone’s not in the room, we can invite them. Bonus:

  • When a “special guest” is no longer needed in a room, they can choose to mute the room from notifications, or leave.

I agree with this regarding “open meetings” in general, though I believe that the current model works against this goal: having many different topics jumbled together in the same room makes mud, not transparency. (It’s a very low signal-to-noise way to accomplish transparency.)

Thanks for the historical reference!

I have to say – I think the reasoning in that thread is not sound. According to that logic, we should just have a single room across the entire project so that everyone can observe everything, because that will increase engagement.

I’m just one human, but I’m pretty sure that’s not how engagement works. The reality is that when the signal-to-noise ratio of an info stream is low, people simply ignore it. And in my experience, increasing that ratio is what drives engagement.

I’ll assume that the elder gods have spoken here and drop this aspiration like so many others. But maybe the fact that this idea comes up regularly is an indicator.

I’ll chime in here with my two cents.

I like the way it is now, I agree it’s better for the “lurking” part, however, I myself sometimes stumble upon a meeting that is having a discussion about something I know, if every team had it’s own special room, that could never happen.
However, for the current setup to work, we need to make Fedocal better - some meetings aren’t even there (infra, for example) and it isn’t integrated with Matrix.

Most of the time, I find it more useful and easier to check the meeting logs and just guess if the meeting is happening at the same time from the older meetings and in my opinion that just makes Fedocal useless.

Like it was written in the older post, something like a !nextmeeting(s) would be useful, and make sure every group has their meetings in Fedocal

3 Likes

I like it the way it is.
I can watch some meetings I’m interested in.

Sometimes the meetings overlap, it is like in a real office where you bump into people and have to vacate rooms. When that happens you break off into an other space like the regualr SIG channel.

:100: to that. Fedocal is a wasteland of outdated information. Meeting room names are duplicated and triplicated between the original rooms on freenode, then libera.chat, and now on Matrix (sometimes multiple ones with slightly different formatting).

For this to become useful again, I think we’d need to 1. clear out and delete “locations” (i.e. meeting rooms) that are no longer used, 2. use consistent / correct formatting for Matrix rooms and remove accidental duplicates, and 3. assign meetings to the rooms they actually already happen in.

Pet peeve: Rename “Fedora Meeting” to “Fedora Meeting 0” to match “Fedora Meeting (1,2,3)”. :grin:

3 Likes

:100:

1 Like

Huh, I was just chasing some links out of curiosity and it looks like about five years ago Matthew said:

It’s from a whole thread that was discussing the possibility of using Discourse’s calendar for some of these things:

An idle thought for the future of Fedora Calendaring