FOSDEM 2020 is over. As I type this, I’m on my way back home from the conference in Brussels. And it has been nice. In the end. I must confess I was frustrated by the middle of the first day, though things got better after that.
Here is the thing: FOSDEM is not your usual conference. There are lots of things going on at once and all of them are crowded. Really, really crowded—to the point where the situation doesn’t make any sense unless you know how to work around it.
Luckily, by the enf of the first day, I met with the NetBSD folks and got some insight on how to navigate the conference. With those tips in mind, Sunday was much more pleasant. So if you are planning on going to FOSDEM in the future, here are those tips and some of my own!
The first thing to internalize is that there are queues for everything. In particular, there are lines to get into the talks. Huh, what?
The way this works is “simple”: each room has an Open/Full sign outside. When the room is at capacity, the sign is flipped to Full and people start queuing. The key thing to realize is that they are not queuing for the current talk: they are queuing for the following talk, and there are good chances that they’ll get in.
When a talk finishes, a fair amount of people usually leave the room and the next batch gets in. But some talks are quite popular, which means not enough people may leave to let the rest of the line in. If the talk sounds like it will be of general interest, be sure there will be a line.
Here is your strategy: if you like most of the talks in a room or want to hang out with a specific community, go there early in the morning and stay there. Do not leave the room. If, instead, you want to see a specific talk only, make your way to the room two talks before so that you can queue and get in for the talk right before the one you want.
The second thing to know is that you should plan ahead. I’m used to just showing up at conferences and deciding “on the fly” whether I want to get into the room for the upcoming talk or not. There usually are one or two rooms after all, and they have sufficient capacity.
That hardly works at FOSDEM. You need a plan. The schedule is so broad and varied (there are possibly tens of concurrent tracks!) that you can’t decide where to go in the split of a second. Make a rough plan of the key talks that you’d like to attend. Double-check that they don’t overlap and check if you’ll have the necessary time for queuing. If not, don’t bother going and find something else to do.
If you are there to meet people instead of watching talks, plan that too. More on this in the next section.
Familiarize yourself with the campus map upfront. There are many different buildings hosting events, so if it’s the first time trying to get somewhere, the “just go to building X and find the room” will take more time than you think due to all corridors being blocked by people.
The third thing to consider is that the topics covered in this conference are extremely varied. I’m used to going to conferences that surround a narrower topic (“BSD”, “Bazel”) so it’s relatively easy to show up and have a common ground to talk with anyone at random.
But at FOSDEM… the topic is “open source”, which is way too broad. I’m sure you can have nice conversations with anyone at random too, but unless you have zero trouble breaking into groups and talking to strangers, you’ll have a bad time. You may end up feeling lonely.
My strategy for next time: if you know somebody, plan ahead when and where to meet them. There are so many people that the usual thought of “walk around and find someone you know when you are bored” doesn’t work either. And because the conference is so varied, the people you know will be spread out in random places. (Yes, I did know people in advance, but because I made bad assumptions about how the conference worked in general, I hadn’t made any plans to meet them either!)
Staying in the city center is a good idea if you want to hang out in the evening. From there, you can take bus 71 to the conference. But… as you can imagine, the bus is very crowded as well. Do not believe the time estimates that Google Maps will give you: they’ll be much longer as the bus will be slower than it should. If you can avoid the “rush to the conference” (that is, 9-10AM), do so.
The bus and trams (I heard this changed since last year) can be paid in cash, but only in coins. At the time of this writing, a ride was 2.50 EUR if paid inside the vehicle, or 2.10 EUR if paid outside on the machine. Note that not all stations have ticket machines.
Applying for a talk
Alright, so why did I go? Well, because I submitted a talk on Bazel and sandboxfs and it got accepted! (See the slides).
However, I low-balled it. When you are applying, you are asked to choose from any of the main rooms, or the lightning talks, or the devrooms. What are these? How do you know where to apply? I had no idea, and as my talk sounded quite niche, I ended up proposing a lightning talk. In retrospect, I would have wanted to do something bigger.
Here is how it works: the “main room” tracks host the primary talks of the conference that are not devroom specific. This makes them sound more important than anything else, but I’m not sure that’s the case. Specifically, talks in the devrooms can be extremely popular too. So what are the devrooms then? Essentially, devrooms are topic-specific mini-conferences. They are not rooms where developers of that topic hang out in a hackathon, as I kinda imagined. They are rooms with their own schedule of talks; just with fewer seats.
I had always been curious about FOSDEM. About 15 years ago, when I was super-active in NetBSD development, I remember other members of the community encouraging me to make the trip up to Brussels… but I kinda let it go because I had never gone abroad on my own and was nervous about it. Meh. I definitely should have gone earlier.
Brussels is a pretty cool city actually, and it feels a lot like Barcelona. This time around I did not see the sun for three days, but the temperature was pleasant.
So. Will I be there next year? Who knows! While I had dinner with my Bazel teammates yesterday evening, we were discussing to propose our own devroom on Build Systems at large, so runing that would be a good excuse. Or maybe I’ll have something new and exciting to talk about, to a larger audience.