Off the Beat: Bruce Byfield's Blog

Last week, ApacheCon hit Vancouver. I duly attended, but a chronic knee problem forced me to sit out most of the conference at home. But I did manage to limp to the bar camp on the Saturday, so I only feel mildly short-changed.

Bar camps, as you may know, began in 2005 as a democratic alternative to O'Reilly Associate's Foo Camp. Short for "friends of O'Reilly," Foo Camp is an invitation-only event. By contrast, bar camps are open to all comers, their name punning on "foobar," the placeholder name used by some coders.

Like Foo Camp, bar camps are explained by describing them as "unconferences." Their rules are simple: at the start of the day, attendees gather around a time table using a white board or post-it notes. Everyone who wants to organize a discussion places the topic on the schedule, teaming up (if necessary) with anyone with a similar topic. People then attend the discussions that interest them, or start breakout sessions to continue interesting conversations. At a successful bar camp, more people can be attending the breakout sessions than the scheduled sessions.

In other words, bar camps are a marginally more organized version of the hall cons that take place at conferences. The difference is, instead of standing around uncomfortably, unable to find enough chairs for everyone in the conversation, bar camps are usually held in places likes work-share offices, where there are plenty of places to sit and usually enough rooms that those who want privacy or quiet can have it.

Alternatively, you might compare bar camps to the evening parties at a technical conference -- the difference being that, at bar camp, nobody is drinking, so you usually get more sense out of them.

Bar camps vs. conferences

Don't get me wrong: I haven't attended a conference yet that I hated or didn't make me feel part of a larger community. Admittedly, I work from home, so I probably amuse easily; at times, taking out the garbage can be a stimulating distraction. Yet despite the occasional unpleasant encounter or event, on the whole I approach a conference like a kid going to the movies, just waiting to be entertained and pleased.

The only trouble is, free software conferences these days are huge. Unless you make arrangements to meet someone, you and your friends can wander the conference for its entirety, and never meet.

Even worse, as happened to me several times at LinuxCon last August, you can catch a glimpse of someone with whom you want to talk -- maybe only the flash of a name tag as the person turns away -- only to find that the person has disappeared into the depths of the crowd and vanished for the duration. Often, you can't even shout after them, because the background noise is too great for them to hear you.

By contrast, while there have been bar camps with thousands, most are under two hundred, and many cap attendance at one hundred or even fifty. Add the fact that the venue is often an open floor plan office, and you rarely have any trouble bumping into anybody you want to meet.

If anything, the problem is just the opposite of the one you face at conferences: If there's somebody you want to avoid, you seem to be standing face to face with them every time you turn around. But, usually, the advantage of a bar camp is that, unless someone is very popular, you can generally have the conversation with them that you want.

Another problem is the kind of exchange that you can have at a conference. When I was an academic-for-hire, I taught largely at Simon Fraser University, which at the time was famous for its tutorial system. Under this system, even first year students met with instructors or teaching assistants in groups of fifteen or less once or twice a week. As the semester continued, and students dropped classes or skipped them for various reasons, any given tutorial might have as few as seven to ten participants.

Having benefited from these tutorials as both student and instructor, I consider this size
as optimal for intellectual exchange. It's a group large enough for a diversity of opinions, yet small enough that only the extremely shy balk at expressing themselves.

The problem is, almost no sessions at a modern free software conference are this size. At times, you might find a tutorial or a bird of a feather session with less than twenty people in attendance, but you can't count on them staying small. Even worse, a regularly scheduled talk can draw several hundred people, a keynote talk several thousands.

This size of an audience means that lectures are the only format that really work for a conference. That means that communication is mostly one-way, from a single person to the many. Even questions at the end rarely alter this exchange, since usually audience members respond to each other through the speaker, not directly. People in the back might not even hear a question until the speaker summarizes it.

This one-way communication isn't a problem when you know nothing about the topic and you've come to learn. But when you want to test your understanding through discussion -- which for me, with the tutorial model embedded in my mind is an absolute necessity (to say nothing of more fun) -- the lecture format has serious limitations. Often at a conference, you can't even talk for more than a few moments afterwards, because another session is starting that you want to hear, or you need to grab a bite to eat before your stomach rebels.

A bar camp, though, is nothing but a series of tutorials. Just about the only requirement at a bar camp is that, if you attend, you have to be prepared to participate. Nobody has to volunteer to run or lead a session, but they do have to be prepared to participate. Instead of remaining passive, as at a lecture, in a bar camp session, you're expected to contribute. The result is a more interesting time for you and everyone else -- and, as a side-effect, a stronger chance of getting to know people who share your interests.

One or the other

If I had my way, talks at a conference would be followed by several breakout sessions, where the audience would divide into small groups and move from being passive recipients of knowledge to active educators of themselves and each other.

Yes, I know that the timing would be difficult to coordinate, and that most conference facilities aren't set up for such a combination. But I can dream. And failing that ideal I wish that there more bar camps and fewer larger conferences.

I don't want to choose between them, but if I had to, give me a bar camp over a conference every time. After a day of passive listening at a conference, I feel enervated, but, after a day of active discussion at a bar camp, I feel ready to take on the world.